Professional Documents
Culture Documents
FALLON 1987 Constructivist Coherence Theory of Constitutional Interpretation
FALLON 1987 Constructivist Coherence Theory of Constitutional Interpretation
APRIL 1987
VOLUME 100
NUMBER 6
C ONSTITUTIONAL
1189
119o
[Vol. Ioo:n189
5 Perry, Interpretivism, Freedom of Expression, and Equal Protection, 42 OHIO ST. L.J.
261, 287 (i98i) (footnote omitted). See generally L. LEVY, LEGACY OF SUPPRESSION (i960).
6 Cf. J. ELY, DEMOCRACY AND DISTRUST: A THEORY OF JUDICIAL REVIEw IO5-6 (1980)
(arguing that "strict review" of restrictions on free speech is not always appropriate when the
government is not trying to stifle communication of a particular message).
1987l
CONSTRUCTIVIST COHERENCE
1191
9 See Bethel School Dist. v. Fraser, io6 S.Ct. 3159, 3165 (1986) (finding that "it is a highly
appropriate function of public school education to prohibit the use of vulgar and offensive terms
in public discourse"); Board of Educ. v. Pico, 457 U.S. 853, 885, 889-91 (1982) (Burger, C.J.,
dissenting).
10 The usefulness of my characterization of this problem as one of "commensurability" does
not depend on the merits of contending positions in debates in the philosophy of science about
or discourses cannot be translated into each other without loss of meaning and disputes cannot
be resolved by reference to shared criteria. See T. KUHN, supra, at 198-204; R. RORTY,
PHILOSOPHY AND THE MIRROR OF NATURE 322-33 (1979). I make limited use of a related
1192
[Vol. lOO:1189
456 (198i) (defining "commensurability" as "the quality or state of being commensurable" and
"commensurable" as "having a common measure").
1 See R. DWORKIN, LAW'S EMPIRE 4-14, 45-86 (1986). Roughly speaking, an "interpretive"
approach to constitutional theory embodies four basic assumptions. First, constitutional interpretation is a "practice" with an internal normative structure. See infra pp. 1232-37. Second,
to participate in the practice requires conformity to the practice's implicit rules or norms. See
infra notes 211-13 and accompanying text. Third, "[viery often, perhaps even typically, the
raw behavioral data of the practice - what people do in what circumstances - will underdetermine" any purely descriptive or positive theory of the practice's implicit norms or the values
or purposes that give the practice its nature. R. DWORKIN, supra, at 52. The data, in other
words, will be consistent with "different and competing" accounts of the practice's rules, nature,
or purposes. Id. Fourth, "[i]f the raw data do not discriminate between these competing
interpretations, each interpreter's choice must reflect his view of which interpretation proposes
the most value for the practice - which one shows it in the better light, all things considered."
Id. at 52-53. For a fuller account, and for an explanation of where I depart from Dworkin as
well as where I follow him, see pages 1231-37 below.
12Although my claims in this Article about the practice of constitutional interpretation will
often be framed in very general terms, several qualifications are necessary. As in most writing
about constitutional theory, the bulk of my arguments and assertions takes as its implicit
paradigm decisionmaking, advocacy, and argumentation in the Supreme Court and in lower
court cases in which there is no Supreme Court precedent closely on point. Although the result
may be a theory that is excessively Supreme Court focused, this approach has advantages. A
theory that takes the Supreme Court as its paradigmatic decisionmaker is able to address directly
19s71
CONSTRUCTIVIST COHERENCE
1193
theory" that I offer has two main aspects. 13 The first asserts that the
implicit norms of our constitutional practice call for a constitutional
interpreter to assess and'reassess the arguments in the various categories in an effort to understand each of the relevant factors as prescribing the same result. The desirability and plausibility of this
strategy, and its implied claim that coherence of this kind is most
often attainable, may seem counterintuitive. But within our legal
culture, it is the rare judicial opinion, the anomalous brief, the unusual
scholarly analysis that describes the relevant kinds of arguments as
pointing in different directions. Typically, legal arguments - including those of judicial and even Supreme Court opinions - find the
best arguments in all of the categories to support, or at least not to
be inconsistent with, a single result. 14 Nor, I shall argue, does this
familiar alignment of arguments reflect sheer intellectual dishonesty
or rhetorical or argumentative excess. The various kinds of constitutional argument are substantially interrelated and interdependent.
Reciprocal influences among them make it possible most of the time
to achieve constructivist coherence. The role of value arguments is
especially important in this respect. I shall be particularly concerned
with showing how value arguments infuse and inform the arguments
that are advanced within other categories.
The second element of my theory comes into play only when the
effort to achieve coherence does not succeed. In such cases, the
categories of argument are assigned a hierarchical order in which the
highest ranked factor clearly requiring an outcome prevails over lower
ranked factors. In particular, I shall argue, the implicit norms of our
the central question of how our system assumes that constitutional issues are and ought to be
settled by the one authority clearly entitled to decide every case on the basis of its own
understanding of highest constitutional principle. Similarly with respect to legal advocacy, a
theory developed around Supreme Court argumentation answers the revealing question of how
our practice assumes that lawyers ought to argue constitutional issues when questions of highest
constitutional principle are most clearly open for decision.
If the Supreme Court focus is distorting, it may be particularly so with respect to the role
played in our constitutional practice by the law school professoriat. Undoubtedly, legal scholars
have an important part in our constitutional practice and, more than judges or practicing
lawyers, they may assume a relatively distinctive and detached role. Even here, however, the
vast bulk of legal scholarship tends to accept and to argue within the implicit and explicit
standards established by Supreme Court practice.
13 For a discussion of the significance of the name "constructivist coherence theory," see note
230 below.
14This is not to say that legal arguments always, or even typically, refer explicitly to all of
the factors that sometimes are recognized as relevant. In some cases, one kind of argument is
emphasized, while others are omitted entirely. Even when this occurs, however, lawyers, judges,
and even Supreme Court justices rarely acknowledge that some relevant factor calls for a result
different from that which is being reached. M'or is it common for dissenting opinions to argue
that the majority has erred by ignoring a relevant factor. Even when factors of general relevance
are not referred to explicitly, it thus seems to be assumed that the conclusions to which they
would point would at least be not inconsistent with the result reached.
1194
[Vol. ioo:ii8 9
I-9ig (i982), and Moore, A Natural Law Theory of Interpretation, 58 S. CAL. L. REV. 279
(1985).
1087]
CONSTRUCTIVIST COHERENCE
1195
1196
[Vol. 1oo:1189
One reason we see relatively few arguments from the text is that
the language of the Constitution, considered as a factor independent
from the other kinds of argument familiar in constitutional debate,
resolves so few hard questions. 21 Nonetheless, arguments from text
can fulfill three functions. Occasionally, an argument from text will
require a unique conclusion - for example, that the President must
be at least thirty-five years old. 2 2 More commonly, arguments from
the text achieve the somewhat weaker but nontrivial result of excluding one or more positions that might be argued for on nontextual
grounds. 23 Thus, although the text of the eighth amendment may not
tell us precisely what "cruel and unusual punishments" are, 24 the
language does require that the amendment's prohibition apply only to
actions that can plausibly be described as "punishments." 2 5 Finally,
among the meanings that are not excluded by arguments from text, a
narrowly text-focused reading will sometimes yield the conclusion that
some are more plausible than others.
In talking about "arguments from text," the "meaning" of the text,
and even its "plain meaning," I do not deny that any attribution of
shall make no law... abridging the freedom of speech." U.S. CONST. amend. I. In deciding
first amendment cases, Justice Hugo Black often emphasized "no law" and argued that this
language established the unconstitutionality of any statute punishing speech. J. ELY, supra note
6, at 1O5; see H. BLACK, supra note 4, at 45. His was an argument from, as well as about,
the text. Justice Harlan, on the other hand, typically espoused a methodology in which he
balanced competing arguments to determine what rights existed. Rejecting Justice Black's
argument from text as unpersuasive, Justice Harlan found that the plain language of the text
left open a range of possible meanings, with the choice among them appropriately determined
by other factors. See, e.g., Konigsberg v. State Bar, 366 U.S. 36, 49 (g61) (rejecting "the view
that freedom of speech and association, as protected by the First and Fourteenth Amendments,
are 'absolutes' . . . in the sense that the scope of that protection must be gathered solely from
a literal reading of the First Amendment" (citation omitted)).
21The first amendment again provides an apt example. Even if read in wholly literal terms,
the language does not forbid abridging all speech, but only "the freedom of speech." See A.
MEIKLEJOHN, POLITICAL FREEDOM: THE CONSTITUTIONAL POWERS OF THE PEOPLE 21 (ig6o).
It is by no means obvious what "the freedom of speech" encompasses. Cf. Konigsberg, 366 U.S.
at 49 & n.io (noting that an "absolutist" view that the first amendment protects all speech
"cannot be reconciled with the law relating to libel, slander, misrepresentation, obscenity,
perjury, false advertising, solicitation of crime, complicity by encouragement, conspiracy, and
the like").
22 But see Peller, supra note ig, at 1174.
23 See Schauer, supra note 17, at 802-12, 824-31. Although I cannot argue the point here,
I acknowledge that certain post-structuralist, deconstructionist writers would regard as contentious the claim that the range of possible meanings of any text is significantly constrained. See,
e.g., Dalton, An Essay in the Deconstruction of Contract Doctrine, 94 YALE L.J. 997, ioo607 (1985); Singer, The Player and the Cards: Nihilism and Legal Theory, 94 YALE L.J. 1, 9-25
(1984).
24 U.S. CONST. amend. VIII.
2S Some incidents cannot be so described. Cf. Bell v. Wolfish, 441 U.S. 520, 535 n.16 (1979)
(relying in part on an argument from the text that the government's infliction of harm could
not be "punishment" in the constitutional sense unless it occurred subsequent to criminal conviction); Ingraham v. Wright, 430 U.S. 651, 664 (1977) (same).
1987]
1197
also assume that the reader, even in assessing arguments about "plain
meaning," will understand that she is reading a constitutional text,
which implies that the language is situated in an interpretive tradition
and must be read with at least a tacit awareness of the range of
extratextual concerns that constitutional interpretation conventionally
takes into account. 27 Nevertheless, the labeling of "arguments from
text" as a distinct category of argument presupposes that it is possible
to read the Constitution with more or less of a narrowly textual
focus.
28
troduction to this literature, see Hoy, Interpreting the Constitution: Hermeneutical and PoststructuralistPerspectives, 58 S. CAL. L. REV. 135 (1985). Important and controversial contributions to the legal literature include R. DWORKIN, cited in note ii above; R. DWORIN, A
MATTER OF PRINCIPLE 146-76 (1985); Brest, Interpretation and Interest, 34 STAN. L. REV.
765, 771 (1982); Cover, The Supreme Court, 1982 Term - Foreword: Nomos and Narrative, 97
HARV. L. REV. 4 (1983); Fish, Fish v. Fiss, 36 STAN. L. REV. 1325 (1984); Fish, Working on
the Chain Gang: Interpretation in Law and Literature, 6o TEX. L. REV. 551 (1982) [hereinafter
Chain Gang]; Fiss, Comment, Conventionalism, 58 S. CAL. L. REv. 177 (1985); Fiss, Objectivity
and Interpretation, 34 STAN. L. REV. 739 (1982); Levinson, Law as Literature, 6o TEX. L.
REV. 373 (1982); and White, Law as Language: Reading Law and Reading Literature, 60 TEX.
L. REv. 415 (1982).
27 See Brest, supra note 18, at 206; Fish, Fiss v. Fish, supra note 26, at x335-36.
28 See Moore, supra note 15, at 320-2I.
29 For a fuller discussion of "originalism" as a fully developed constitutional theory, see pages
1211-14 below.
30 See, e.g., South Carolina v. United States, 199 U.S. 437, 448 (1905) ("The Constitution
is a written instrument. As such, its meaning does not alter. That which it meant when adopted
it means now."); T. COOLEY, A TREATISE ON THE CONSTITUTIONAL LIMITATIONS WHICH REST
UPON THE LEGISLATIVE POWER OF THE STATES OF THE AMERICAN UNION 55 (Da Capo Press
ed. 1972) (ist ed. 1868) ("The meaning of the Constitution is fixed when it is adopted, and it is
not different at any subsequent time when a court has occasion to pass upon it."); Brest, supra
note IS, at 208-09; Monaghan, Our Perfect Constitution, 56 N.Y.U. L. REv. 353, 373-77
(i98i); Munzer & Nickel, Does the Constitution Mean What it Has Always Meant?, 77 COLUM.
L. REV. 1029, 1043-44 (1977).
1198
[Vol. 1OO:1I8 9
porary meanings. 3 1 In developing a typology of constitutional argument, it is enough to recognize that both types of arguments from the
text can and do get made. Consideration of how these competing
claims get resolved must await later parts of this Article.
B. Arguments About the Framers' Intent
Searches for the meaning of a constitutional provision frequently
involve inquiries into the intent of the framers and ratifiers. 3 2 Controversy abounds concerning the weight that intent ought to have.
Although "interpretivists" view the intent of the framers as controlling, 33 most other constitutional lawyers regard intent as entitled to
only some, not very clearly specified, weight. 34 Moreover, several
important scholars have recently argued that the intent of the framers
generally has no justifiable place in constitutional argument. 35 But
this form of nonintentionalism is more plausibly viewed as a prescriptive proposal than as an account of existing practice. It is relatively
uncontroversial that the Supreme Court regards the framers' intent as
36
an important factor in constitutional adjudication.
Notoriously, searches for intent divide into several types. 3 7 One
helpful division distinguishes between "specific" or "concrete" and
31 See P. BOBBITT, supra note I, at 25-26; Perry, supra note 17, at 564-65.
32 See, e.g., Ford v. Wainwright, io6 S. Ct. 2595, 2600-02 (1986); Marsh v. Chambers, 463
U.S. 783, 788 (x983); P. BOBBITT, supra note 15, at 9-24; Monaghan, supra note 30, at 37579.
33See infra pp. 1209-17.
34 See, e.g., Brest, supra note 18, at 224, 229-34; Perry, supra note 17, at 569-70.
35 See R. DwoRKIN, supra note ii, at 359-81; Moore, supra note I5, at 338-58; Simon,
The Authority of the Framers of the Constitution: Can OriginalistInterpretationBe Justified?,
73 CALIF. L. REV. 1482 (1985); see also Sandalow, ConstitutionalInterpretation, 79 MICH. L.
REv. 1033, io62-64 (i981) (arguing that although there has been little incentive to develop an
interpretive framework justifying departures from the framers' intentions, it is mistaken to think
that the framers' intentions and expectations furnish even a "core" of constitutional meaning
that must be respected).
36 See cases cited supra note 32. To deny this assertion, it would be necessary to adopt a
thoroughly "realist" attitude toward arguments from the framers' intent. On this interpretation,
even though the Supreme Court frequently talks about the intent of the framers, the Court
manipulates its findings, or adjusts the level of abstraction at which the framers' intent is
specified, see infra text accompanying notes 37-41, in order to rationalize results reached on
other grounds. I certainly have no quarrel with the common observation that the Supreme
Court frequently relies on poorly executed historical scholarship to identify the intent of the
framers. See Kelly, Clio and the Court: An Illicit Love Affair, x965 SuP. CT. REv. ixg. Nor
would I question that, when the Court does so, result-orientation frequently provides the most
plausible explanation. But the realist interpretation of the role of the framers' intent is ultimately
unpersuasive, because it mischaracterizes the phenomenology of constitutional argument and
interpretation: it omits the sense of the judge or constitutional lawyer that, as she searches the
legislative history, it does and ought to matter what she finds there.
37 See, e.g., Brest, supra note 12, at 223; Speech by Judge Robert H. Bork (Nov. x8, 1985),
reprinted in THE GREAT DEBATE: INTERPRETING OUR WRITTEN CONSTITUTION 43, 47-48
19871
CONSTRUCTIVIST COHERENCE
1I99
(Federalist Society 1986). The most prolific expositor of distinctions has been Professor Dworkin.
See, e.g., R. DWORKIN, TAKING RIGHTS SERIOUSLY 134-37 (1977) (distinguishing the "concepts"
that the framers incorporated into the Constitution from their "conception" of what, in particular,
those concepts require); R. DWORKIN, A MATTER OF PRINCIPLE, supra note 26, at 44 (distinguishing the framers' interpretive hopes from their interpretive expectations).
38 See Perry, supra note 3, at 287.
39 See R. BERGER, GOVERNMENT BY JUDICIARY 117-33 (1977); Bickel, The Original Understanding and the Segregation Decision, 69 HARv. L. REv. 1, 58-59 (1955). Even Berger's critics
generally have not disputed this point. Perry, supra note 3, at 292 n.131; see Soifer, Protecting
Civil Rights: A Critique of Raoul Berger's History, 54 N.Y.U. L. REv. 651, 705 (I979).
40 See generally Bickel, supra note 39.
41 See Brown v. Board of Educ., 347 U.S. 483 (i954). The situation would of course be
different and more complicated if the framers had specifically and explicitly intended that the
equal protection clause not under any circumstances be construed to require desegregated schools.
But the problem of "negative intent" was not presented by Brown. See infra text accompanying
notes 390-94.
Paralleling the distinction between specific and general intent is Professor Dworkin's distinction between "concepts" and "conceptions." See R. DWORKIN, TAKING RIGHTS SERIOUSLY,
supra note 37, at 134-37. Within this terminology, the authors of the fourteenth amendment
intended to constitutionalize a requirement of equal legal treatment. This concept, however,
admits of many conceptions or interpretations. While the framers' own conception may have
been a narrow one - possibly permitting, for example, segregation in the schools - the concept
of equal treatment also may have more exacting conceptions. Dworkin argues that it is the
concept of equal treatment, not the narrow conception of the framers, that the fourteenth
amendment properly expresses.
1200
[Vol. 1oo:1189
1987]
CONSTRUCTIVIST COHERENCE
1201
were under the Constitution, it would make no sense, Marshall reasoned, for the states to be able to frustrate constitutionally legitimate
federal policies. 48 Arguments of this kind can be viewed as ones of
constitutional theory because, although they do not rely on either the
precise linguistic meaning of particular constitutional provisions or on
the historically identifiable intent of the framers, they are text focused.
Finally, closer to the particularistic end of the theoretical spectrum
lie such assertions as the commonplace that it is the purpose of the
first amendment to protect a marketplace of ideas. 49 It would be
possible, of course, to cast this claim as an account of the framers'
general intent. But the argument that courts should protect the marketplace of ideas might also issue, entirely independently of any inquiry into the attitudes of the framers, from a view about the values,
assumptions, or political theory in light5 0of which the language of the
first amendment makes the most sense.
Arguments of this kind are closely analogous to, if not identical
with, what have sometimes been characterized as "clause-bound" interpretivist arguments.3 1 Clause-bound interpretivist arguments purport to "derive" a value from some relatively clear and secure foundation in a particular provision of the constitutional text; the
arguments are classified as interpretivist because the aspiration is to
draw "inferences" purely from textual sources. 5 2 But the terminology
of "derivation" and "inference" is misleading. At least after we have
left the domain of arguments from text, it is always necessary to
formulate a theory about a constitutional provision, or ascribe a purpose to it, before any "derivation" of particular conclusions can occur.
Moreover, because the text commonly is so spare, it frequently will
fail to determine any one theory but will, instead, be consistent with
several. Thus the first amendment appears to some to create a general
marketplace of ideas. But others ascribe to the free speech guarantee
48 Chief Justice Marshall also relied importantly on structural arguments in his even more
famous opinion in Marbury v. Madison, 5 U.S. (i Cranch) 137 (1803), in which he reasoned,
for example, that a written constitution would be meaningless in restraining legislative power,
as it was clearly intended to do, if the courts were incompetent to engage in judicial review.
See id. at 178.
49 See, e.g., Red Lion Broadcasting Co. v. FCC, 395 U.S. 367, 390 (1969); Abrams v. United
States, 250 U.S. 616, 630 (i919) (Holmes, J., dissenting).
so It is for this reason that John Stuart Mill, who authored a celebrated essay on the values
and purposes served by preserving wide liberties of thought and discussion, is frequently
classified as a first amendment theorist. See, e.g., Wellington, On Freedom of Expression, 88
YALE L.J. 1105 (1979).
the first amendment or its framers, his arguments establish a framework of values and purposes
that can usefully be referred to in defining the "freedom of speech" that the first amendment
should be construed to protect.
51 See J. ELY, supra note 6, at 11-41.
52 See, e.g., Bork, Neutral Principles and Some First Amendment Problems, 47 IND. L.J.
1, 3-7 (971).
1202
[Vol. ioo:ii8 9
the purpose, for the most part, of protecting the value only of indi53
vidual - as opposed, for example, to corporate - self-expression.
Still others view the object as limited to preserving the freedom of
54
political speech essential to a democracy.
This disparity of views should not be surprising. A constitutional
theory must seek not only to explain constitutional guarantees and
prohibitions but also to do so in a normatively attractive way. And
where more than one theory plausibly accounts for the text having
been written as it was, an assessment along a normative dimension,
whether undertaken consciously or unconsciously, becomes inevitable
and desirable. The upshot is that the "derivation" of constitutional
values can seldom if ever be a value-neutral enterprise.5 5 That is
why what others sometimes call clause-bound interpretivist arguments
are more accurately classified as arguments of constitutional theory.
D. Arguments from Precedent
Constitutional disputes frequently abound with analysis of the
meanings of judicial precedents. 5 6 Indeed, constitutional arguments
sometimes address themselves almost entirely to the meanings of previously decided cases:5 7 read one way, precedent indicates one result
in dispute, whereas if read another, it leads to a different conclusion.
More commonly, however, prior judicial decisions form a patchwork
into which a current problem must be fitted through a combination
of analytical, analogical, and theoretical reasoning.5 8 The problem
then is one of developing a theoretical account of what the decided
cases stand for.
As Professor Shapiro has recently observed, our legal system supports "two coexisting doctrines of precedent: a narrow one for getting
s3 See Pacific Gas & Elec. Co. v. Public Util. Comm'n, io6 S. Ct. 903, 917 (1986) (Rehnquist,
J., dissenting); see also Baker, Commercial Speech: A Problem in the Theory of Freedom, 62
IOWA L. REv. 1, 4 (1976) (arguing that the first amendment protects only speech connected with
individual liberty and self-realization).
54See Bork, supra note 52, at 26-35.
the past." M. PERRY, THE CONSTITUTION, THE COURTS, AND HUMAN RIGHTS IOii (x982) (explaining, although not endorsing, the aspirations of interpretivism).
56 For interesting and useful accounts of the role of precedent in constitutional adjudication,
see, for example, Michelman, Constancy to an Ideal Object, 56 N.Y.U. L. REV. 406 (I981);
Monaghan, supra note 30, at 387-91; Monaghan, Taking Supreme Court Opinions Seriously, 39
MD. L. REv. 1 (1979) [hereinafter Court Opinions].
57See Jones, supra note 18, at 28; see, e.g., Skipper v. South Carolina, io6 S. Ct. 1669,
1670-73 (1986) (interpreting two earlier decisions to elucidate the standard of admissibility of
mitigating evidence in a criminal case); Michigan v. Jackson, io6 S. Ct. 1404, 1407-1I (1986)
(discussing earlier cases on the admissibility of confessions in criminal cases).
58 See generally E. LEVI, AN INTRODUCTION TO LEGAL REASONING (1949).
1987]
CONSTRUCTIVIST COHERENCE
1203
59 Shapiro, In Defense of Judicial Candor, ioo HARV. L. REv. 731, 734 (987). Shapiro
attributes this view, which he describes as being "close" to his own, to K. LLEWELLYN, THE
BRAMBLE BUSH 74-75 (1951).
60 465 U.S. 89 (1984).
61 Id. at 165 (Stevens, J., dissenting). Joining Justice Stevens' opinion were Justices Brennan,
Marshall, and Blackmun.
62 Id. at o9.
63 For an insightful assessment of the competing arguments, in Pennhurst, see Shapiro, The
Supreme Court: 1984 Term - Comment: Wrong Turns: The Eleventh Amendment and the
Pennhurst Case, 98 HARv. L. REV. 6x (1984).
64 Modern philosophy of science holds that none of our beliefs about the world is independent
of theory. See, e.g., H. PUTNAM, REASON, TRUTH AND HISTORY 135 (198); IV. QUINE, Two
Dogmas of Empiricism, in FROM A LOGICAL POINT OF VIEW 20-46 (2d ed. 198o). The
implication is that the data to be explained along the dimension of "fit" are not independent of
theory, because theory is needed to establish what will count as data and what will not.
65 Shapiro, supra note 59, at 734. Thus, in Pennhurst, either Justice Powell or Justice
Stevens may have been mistaken in his judgment about what readings the narrower and broader
doctrines of precedent would permit.
1204
[Vol. lOO:1189
determining whether restrictions on abortion offend the Constitution "is to resolve the issue by
constitutional measurement, free of emotion and of predilection").
70 For a survey, see Wiseman, The New Supreme Court Commentators: The Principled, the
Political, and the Philosophical, io HASTINGS CONST. L.Q. 315 (1983).
value-laden approach, see Monaghan, cited in note 3o above, at 375-87, and Van Alstyne,
Interpreting This Constitution: The Unhelpful Contributions of Special Theories of Judicial
Review, 35 U. FLA. L. REV. 209 (1983).
19871
CONSTRUCTIVIST COHERENCE
1205
groups or of individuals and those that appeal instead to goals or policies of the society that do
not give rise to rights. Cf. R. DWORKIN, TAKING RIGHTS SERIOUSLY, supra note 37, at 82
(distinguishing "arguments of policy," which "justify a political decision by showing that [it]
advances or protects some collective goal of the community as a whole," from "arguments of
principle," which "justify a political decision by showing that the decision respects or secures
some individual or group right"). Although I shall generally use the term "policy arguments" to
refer to the former and "moral arguments" to refer to the latter, my purposes in this Article
require no sharp distinction between the two.
72 U.S. CONST. amend. V.; id. amend. XIV.
73Id. amend. XIV, I.
74 Id. amend. IV.
75Id. amend. VIII.
76 See Gallie, Essentially Contested Concepts, 56 PROC. OF THE ARISTOTELIAN SOC'Y 167
(1956); MacIntyre, The Essential Contestability of Some Social Concepts, 84 ETHICS I (1973).
77To avoid reliance on the moral and political judgment of judges, one conceivable position
12o6
[Vol. ioo:ii8 9
would be that judges, in construing these provisions, should never give weight to value arguments about what is, for example, fair or unfair, reasonable or unreasonable; instead, they
should base their decisions in every case either on the framers' views or on the value scheme
that generally obtained at the time of the provision's enactment. See Monaghan, supra note 30,
at 367 ("No relevant evidence on the ninth amendment and very little with respect to the
privileges and immunities clause supports an inference that either was intended to have a
dynamic character."). But that view has not prevailed.
78 Ford v. Wainwright, io6 S. Ct. 2595, 2600 (1986) (quoting Trop v. Dulles, 356 U.S. 86,
ioi (1958) (plurality opinion)).
79 See Hudson v. Palmer, 468 U.S. 5,7, 525-28 (1984); see also New Jersey v. T.L.O., 469
U.S. 325, 337 (i985) (stating that courts must identify "the individual's legitimate expectations
of privacy and personal security" and weigh them against "the government's need for effective
methods to deal with breaches of public order").
80 Mathews v. Eldridge, 424 U.S. 319, 348 (1976).
81 For an insightful critique of the Court's approach to assessment of the values actually at
stake in procedural due process cases, see Mashaw, The Supreme Court's Due Process Calculus
for Administrative Adjudication in Mathews v. Eldridge, 44 U. CHI. L. REv. 28 (1976).
A more controversial example of judicial reliance on value arguments emerges from "substantive due process" cases. Substantive due process doctrine assumes that the due process
clause protects at least some substantive values deemed to be "fundamental." See, e.g., Roe v.
Wade, 410 U.S. 113, 152-55 (i973); Poe v. Ullman, 367 U.S. 497, 541 (196i) (Harlan, J.,
dissenting). But substantive due process methodology also postulates that the protected values
must not only be given content, but actually identified, by appeal to some authority outside of
the explicit constitutional text. In Moore v. City of East Cleveland, 431 U.S. 494 (i977), for
1987]
CONSTRUCTIVIST COHERENCE
1207
In a second kind of case, the invocation of value arguments appears to reflect the slightly different assumption that, where arguments
within other categories are indeterminate or closely balanced, it is
either necessary or at least reasonable to take account of considerations
that value arguments address. The claim here, though difficult to
document, resonates with both experience and common sense. A clear
example, already discussed at some length, lies within the category of
arguments from precedent. The implicit norms of our constitutional
practice frequently permit both broad and narrow readings of past
judicial decisions. 82 Which approach ought to be preferred in any
particular case, and how a string of decisions ought to be ordered into
a pattern or subsumed under a theory, often will and should be
resolved on normative grounds. Similar results occur within other
categories of argument, such as arguments of constitutional theory.
Confronted with contending theoretical arguments that are equally or
nearly equally plausible, judges prefer those that accord with their
83
views of justice or sound policy.
Although I have provisionally distinguished between two kinds of
value arguments - those needed to give meaning to constitutional
provisions that expressly require value judgments and those invoked
in making choices where other kinds of arguments are closely balanced
or indeterminate - this distinction is not one that I shall try to
maintain. There are too many unclear cases. Within the Bill of
Rights, for example, the protection of "the freedom of speech"8 4 arexample, the plurality reasoned that the due process clause protects certain family relationships
because "the institution of the family is deeply rooted in this Nation's history and tradition."
Id. at 503. This argument presupposes that the Constitution authorizes the protection of
nonenumerated rights. It does not claim, however, that the constitutional text expressly, or
even within the terms of a constructivist theoretical interpretation, marks family relationships
as deserving of protection or indicates how much protection they ought to get. Rather, the
argument assumes that protection is warranted because history and tradition provide family
relationships with the pedigree necessary for them to deserve weight in constitutional argument.
Cf. Poe, 367 U.S. at 542 (Harlan, J., dissenting) (arguing that in giving content to the due
process clause, the Court cannot rely on any "formula" or "code" but must refer to "the traditions
from which [the country] developed as well as the traditions from which it broke").
82 See supra text accompanying note 59.
83 Value arguments of this type are frequent in first amendment jurisprudence. See Wellington, Common Law Rules and ConstitutionalDouble Standards, 83 YALE L.J. 221, 267-70
0973). For instance, in Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council,
Inc., 425 U.S. 748, 763 (976), the leading case establishing that the first amendment protects
commercial advertising, the Supreme Court argued that the "interest in the free flow of commercial information" was frequently more important to consumers than their interest in "the
day's most urgent political debate" and that courts could properly take this consumer interest
into account. Although this argument could be characterized as one of constitutional theory, it
appears to have played a different role in Justice Blackmun's opinion. With plausible arguments
of constitutional theory available on both sides, the Court appealed to the value of consumer
satisfaction - one that is not clearly established as a "constitutional value" by provisions of the
constitutional text - to determine how the balance should be struck.
64 U.S. CONST. amend. I.
12o8
[Vol. 1OO:1189
guably constitutionalizes a value judgment that is as essentially contestable as the prohibitions of "cruel and unusual" punishments and
"unreasonable" searches and seizures. To determine whether the utterances of a high school student are constitutionally protected, the
Supreme Court has asserted, "[t]he undoubted freedom to advocate
unpopular and controversial views in schools and classrooms must be
balanced against the society's countervailing interest in teaching students the boundaries of socially appropriate behaviour."8 5 Is it accurate in this case to say that the concept of "freedom of speech"
requires the interpreter to resort to sources of value outside the constitutional text to determine how the constitutional guarantee ought
to be construed? Or should we say instead that a constitutional
interpreter may be guided by considerations of value in resolving a
case in which other factors - such as arguments from text, the intent
of the framers, constitutional theory, and precedent - are nearly in
equipoise or at least would be viewed differently by people whose
values are different? All that seems clear is that value arguments will
somehow enter the judicial calculus.
Within the category of arguments of value, a final distinction will
prove helpful. It involves the sources of values to which a judge
might appeal. One kind of value argument refers to some repository
of values, outside of herself, that a judge or lawyer believes to be a
legitimate source of authority in constitutional interpretation. 8 6 That
source might be traditional morality,8 7 consensus values, 88 natural
law,8 9 economic efficiency, 90 or the original position liberal methodology of John Rawls. 91 Another imaginable kind of value argument
85 Bethel School Dist. No. 403 v. Fraser, io6 S. Ct. 3159, 3164 (1986).
86 See generally Bennett, Objectivity in ConstitutionalLaw, 132 U. PA. L. RaV. 445, 447
(1984) (defining objectivity in constitutional law in terms of "sources for decision external to the
decider's own .. .standards or values" even if those external sources are not authoritative in
the sense of determining one right answer to a constitutional question).
87 See, e.g., Moore v. City of East Cleveland, 431 U.S. 494, 503 (1977) (plurality opinion)
("[T]he Constitution protects the sanctity of the family precisely because the institution of the
family is deeply rooted in this Nation's history and tradition."); Poe v. Ullman, 367 U.S. 497,
542 (i96i) (Harlan, J., dissenting) ("The balance of which I speak is the balance struck by this
country, having regard to what history teaches are the traditions from which it developed as
well as the traditions from which it broke."); Wolf v. Colorado, 338 U.S. 25, 28-30 (i949)
(examiming traditions of the "English speaking world" to determine views on the exclusionary
rule).
8sSee Wellington, supra note 83, at 284.
89 See Dworkin, "Natural" Law Revisited, 34 U. FLA. L. REv. I65 (1982); Moore, supra
note I5.
90 See generally R. POSNER, ECONOMIC ANALYSIS OF LAWv (3d ed. 1986).
91 J. RAWLS, A THEORY OF JUSTICE (1971); see Michelman, In Pursuit of Constitutional
Welfare Rights: One View of Rawls' Theory of Justice, 121 U. PA. L. REv. 962 (1973); Richards,
Free Speech and Obscenity Law: Toward a Moral Theory of the First Amendment, 123 U. PA.
L. REV. 45, 59-70 (i974).
CONSTRUCTIVIST COHERENCE
1987]
1209
H-. THE
121O
school is called, attempts to exclude value arguments insofar as possible from the constitutional calculus. The logic of the interpretivist
position calls for reducing arguments from precedent to prior interpretations of text and the framers' intent. 94 Interpretivism also rejects
arguments of constitutional theory that depart from the intent of the
95
framers.
J., dissenting) ("I fear to see the consequences of the Court's practice of substituting its own
concepts of decency and fundamental justice for the language of the Bill of Rights."); id. at 6975, 89-92.
9s This theory of the Constitution's role and status traces to the foundational decision of
judicial review, Marbury v. Madison, 5 U.S. (I Cranch) 137 (1803). See Brest, supra note i8;
Grey, supra note 18, at 705.
99 As will become clear later, this is not an argument that I accept, although I believe that
it must be taken seriously and that its attractiveness helps to explain the power of interpretivism.
The argument's crucial gap involves the implicit assumption that judicial conformity to the
Constitution necessarily requires that judicial decisions be "determined" in every case by some-
1987]
1211
thing other than the reasoned, and publicly explained and defended, value judgments of the
interpreting judges. For an insightful discussion of this and related issues, see Michelman,
Justification (and Justifiability) of Law in a ContradictoryWorld, in 28 NOMOS: JUSTIFICATION
71 (J. Pennock & J. Chapman eds. 1986). For a fuller discussion of the fallacies of interpretivism's implicit theory of the rule of law, see note 13x below.
100 See supra text accompanying notes 29-31.
101 See supra text accompanying notes 37-41. It is a minor irony of theories emphasizing
the intent of the framers that the framers themselves apparently thought that their "psychological
intent" - as distinct from the natural import of their words - was a concept of little if any
validity in constitutional adjudication. See Powell, The Original Understanding of Original
Intent, 98 HARV. L. REv. 885, 887-88, 902-13 (1985). To the extent that intent was useful at
all, many of the framers argued that the relevant intent would be not their own but that of the
ratifiers. See id. at 888, 9o6-07.
102 This is the view of Professor Berger. See R. BERGER, supra note 39, at 283-418.
It
also appears to be one to which Professor Monaghan is attracted, see Monaghan, supra note
30, at 374-81, though less clearly committed, see id. at 382.
103 See Brest, supra note 18, at 223-24, 231-34.
104 See Monaghan, supra note 30, at 378 (arguing that conceptualizing original intent at high
levels of abstraction, as moderate interpretivists do, "sterilize[s] the concept ... [and,] in effect,
removes it as an interpretational constraint").
105 See R. BERGER, supra note 39, at 363-72.
1212
[Vol. io0:II8 9
encounters a daunting array of historiographical, conceptual, and interpretive problems. 10 6 If, for example, the framers are taken individually, a problem exists in defining what, among a person's mental
attitudes, ought to count as an intention. 10 7 Moreover, the originalist
confronts a perplexity that borders on paradox if it turns out that
many of the framers had a certain kind of "interpretive intent": if they
intended that a constitutional provision be adapted judicially to accommodate evolving needs rather than applied only to an historically
defined set of situations.10 8
Once intentions are identified at the individual level, originalists
confront the equally formidable challenge of combining them into a
plausible concept of group intent. Different concerns may have motivated different drafters or supporters even of a single constitutional
provision. For any single person or group, the animating principle
may be stated more or less broadly. Probably as a result, no widely
recognized legal convention establishes precisely how the required
summing of individual intentions ought to occur.10 9
Although it is possible to develop an intelligible conception of
group intent, the best available theory fails to satisfy the aspirations
of originalism. Far from being a simple fact awaiting discovery by
the industrious researcher, the framers' intent must be viewed as an
intellectual construct, developed through a process of interpretation,
that seeks to embody the principles that furnish the best political
justification for a constitutional provision and that find substantial
support in the political climate surrounding the provision's framing
106 See, e.g., R. DWORKIN, A MATTER OF PRINCIPLE, supra note 26, at 38-57; Brest, supra
note i8, at 221-23.
107 A person voting to accept a constitutional provision or amendment may have given little
thought to its meaning. Others may have had "hopes" about how the provision would be
interpreted that differed from their interpretive "expectations." See R. DWORKIN, A MATTER
OF PRINCIPLE, supra note 26, at 44-45; Moore, The Semantics of Judging, 54 S. CAL. L. REV.
151, 265-70 (i98I).
108 See Brest, supra note 18, at 212, 215-16; R. DWORKIN, A MATTER OF PRINCIPLE, supra
note 26, at 51-55. Professor Dworkin has argued that the framers had this interpretive intent
when they chose the broad, open-ended language characteristic of, for example, many of the
guarantees of the Bill of Rights and of the fourteenth amendment. See R. DWORKIN, TAKING
RIGHTS SERIOUSLY, supra note 37, at 132-37; see also Bickel, supra note 39, at 58-65 (advancing
the "hypothesis" that the framers of the fourteenth amendment chose broad language in order
to permit a process of interpretive growth that they did not specifically mandate). Likewise,
the Supreme Court has sometimes assumed that the framers intended to license judicial adaptation. See Weems v. United States, 217 U.S. 349, 372-73 (9qo) (holding that the cruel and
unusual punishment clause of the eighth amendment is not restricted in application to practices
current at the time of adoption); see also Harper v. Virginia Bd. of Elections, 383 U.S. 663,
669 (1966) ("[T]he Equal Protection Clause is not shackled to the political theory of a particular
era."); Home Bldg. & Loan Ass'n v. Blaisdell, 290 U.S. 398, 442-44 (1934) ("It is no an-
swer ... to insist that what the provision of the Constitution meant to the vision of that day
it must mean to the vision of our time.").
109 See R. DWORKIN, A MATTER OF PRINCIPLE, supra note 26, at 38-57.
1987]
CONSTRUCTIVIST COHERENCE
1213
and adoption. 1 10 With the notion so conceived, particular interpretations of the framers' group intent are dependent upon and constrained by historical materials. But they also embody implicit or
explicit normative judgments. Frequently the historical evidence will
fit at least tolerably well with more than one account of the framers'
In this situation, choice among the historianimating intentions.'
cally plausible competitors - in a context in which the purpose of
the inquiry is to guide a judicial decision - is appropriately if not
necessarily made on normative grounds.
This conception of the framers' group intent, of which I myself
shall make use later in this essay, does not imply that the task of
interpretation cannot be judged as being executed well or poorly,
successfully or unsuccessfully. But any conception of the framers'
intent is nevertheless theory dependent and contestable in a way that
undermines the originalist's project of "simply" enforcing judgments
constitutionalized by the framers. In permitting a political evaluation
of contending understandings that find more or less equal support in
the evidence, an "interpretive" conception of the framers' intent affords
arguments of value too much weight to satisfy a consistent originalist.
Nor are originalism's deficiencies limited to its inability to meet its
own standards. Assessed as a descriptive theory of contemporary
constitutional interpretation, originalism fails spectacularly. Originalism cannot account for much of our constitutional practice of at least
the last 50 years. 1 12 Most of our free speech jurisprudence could
probably not be justified within an originalist methodology.11 3 Modern equal protection doctrine, including the landmark decision in
Brown v. Board of Education," 4 affords a series of telling counterexamples to originalism's descriptive claims. So does fourth amendment law. 1 15 Finally, originalism cannot adequately explain the familiar and important role of precedent in constitutional argument.
Strict originalism does not fare much better when appraised as a
normative theory. The Constitution was written to endure through
110 See id. at 326-29. Although Dworkin was concerned with legislative intent as a concept
in statutory interpretation, his idea is easily adapted to the constitutional context. Dworkin,
however, has argued recently that the intent of the framers is not a useful concept in constitutional interpretation. See R. DWORKIN, supra note ii, at 359-69. For other views similar to
those expressed in the text about how the concept of intent is most intelligibly and appropriately
understood, see Moore, cited in note 15 above, at 349, and White, cited in note 26 above, at
438-41.
I See, e.g., infra text accompanying notes 331-35 (discussing the intent of the framers of
204,
York, 445 U.S. 573, 591 n.33 (198o); Katz v. United States, 389 U.S. 347, 352-53 (1967).
121 4
[Vol. IOO:1IS9
different historical ages," 6 and part of the task of constitutional interpretation is to produce a body of law adequate to the present day. 117
The familiar metaphor of a "living Constitution"' 18 suggests that our
legal culture assumes a close connection between legal interpretation
in general, and constitutional interpretation in particular, and an
evolving ideal of justice." 9 In other words, it is, within our tradition,
an important aim of constitutional adjudication to achieve a body of
law that is responsive to perceived modern needs and accepted by the
population as tolerably just. The questions of when, how, and to
what extent the rule-of-law ideals associated with a written constitution should accommodate needs for adaptation and change are notoriously difficult. Yet on a spectrum running from theories that would
allow no adaptation at all to those that would reduce constitutional
law to an unmediated clash of value arguments, originalism occupies
the first untenable extreme. It is a normative deficiency of originalism
that its premises deny any scope whatsoever for accommodation. 12 0
116 See McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 415 (1819) (noting that the Con-
stitution was "intended to endure for ages to come, and, consequently, to be adapted to the
various crises of human affairs").
117 Against a functional critique of this kind, an interpretivist might argue that although
constitutional law has a function, its capacity to fulfill that function depends not on its malleability to accommodate perceived needs but instead on a set of "rule-of-law" virtues associated
with constancy, predictability, and democratic lawmaking. See R. BERGER, supra note 39, at
283-99. But this argument, if offered as a defense of originalism against a charge of functional
misguidedness, has three flaws. First, as I have argued at length, originalism itself is functionally
unworkable; it cannot satisfy the standards that it erects. See supra pp. 1211-13. Comparative
functional assessments are therefore deeply embarrassing to the originalist. Second, the rule-oflaw argument assumes greater determinacy than in fact exists in our constitutional jurisprudence.
Purposive interpretation has no viable formalist alternative. Third, the functional critique of
originalism stated in the text is not purely instrumentalist in its assumptions. An interpretive
approach that treats it as an aim of constitutional interpretation to achieve a functionally
workable and morally just body of law by no means necessarily implies that this is the sole end
of constitutional interpretation or that "rule of law virtues" must be discarded entirely. See
Moore, supra note I5, at 313 (describing "rule-of-law virtues" as "those values that mandate
that judges should not dispense justice in some ad hoc, case-by-case basis"). Text, history, and
precedent both do and ought to impose limitations on how far constitutional provisions drafted
in one era may be ascribed new meanings in another.
I's See, e.g., H. McBAIN, THE LIVING CONSTITUTION (1927); Reich, Tie Living Constitution and the Court's Role, in HUGO BLACK AND THE SUPREME COURT 133 (S. Strickland ed.
1967).
119 See Dworkin, supra note 89; Grey, supra note 18, at 715-17.
120 Sensitive to this difficulty, some originalists have hinted at the possibility of isolated
relaxations. In particular, both Raoul Berger and Professor Henry Monaghan have conceded
that, with nonoriginalist doctrines so thoroughly ensconced in our constitutional law, it would
be unthinkable to overturn them all. See R. BERGER, supra note 39, at 413; Monaghan, supra
note 30, at 382. Professor Monaghan has therefore suggested that it would be possible to reach
an accommodation of sorts by allowing arguments from precedent to prevail "where settled
expectations of the body politic have clustered around" decided cases. Monaghan, Court Opinions, supra note 56, at 7; accord Monaghan, supra note 30, at 382. The problem with this
approach is that it attempts to purchase descriptive plausibility at a price that includes consid-
1987]
CONSTRUCTIVIST COHERENCE
1215
Because it accounts more accurately for our most recent halfcentury of constitutional law, moderate interpretivism might appear
a more attractive alternative than originalism. But moderate interpretivism also fails to provide a viable theory. 12 1 Moderate interpretivism recognizes the legitimacy of arguments of original as well as
122
contemporary meaning and of specific as well as general intent.
But this recognition raises the problem of how to decide which meaning or which type of intent should govern in any particular situation.
Within the category of the framers' intent, for example, arguments
concerning specific intent and those involving general intent embody
different conceptions of what arguments within the category ought to
be about. If arguments of specific intent are more appropriate, then
the best argument of specific intent should always prevail; there is no
basis, internal to the category, on which to say that arguments of
general intent should ever be accepted instead. If, on the other hand,
the category more properly refers to the framers' general intent, then
arguments of general intent ought to win out in every case.
Because there is no common denominator to which arguments of
general and specific intent can be reduced, it is impossible, once the
permissibility of both kinds of argument is accepted, to resolve the
competition between the different conceptions by reference to some
standard internal to the category of arguments of intent. Some outside
arbiter - a standard of value or principle of selection external to the
category - must determine the choice. Yet if reference to outside
considerations must take place, then moderate interpretivism fails in
its ambition for itself, which is to reduce constitutional interpretation
to the dictates of one or two preferred factors that require no appeal
12 3
to other sources of value.
erable sacrifice of intellectual coherence. Monaghan's proposal to cede this much force to stare
decisis accords badly with the background originalist premises that he also purports to champion.
How, in this context, would an originalist deal with earlier decisions that she regards as
illegitimate in their foundation? The distinction between application and extension may have
initial appeal. But, at least at its borders, the distinction is a blurry one; it is a familiar common
law pattern to see lines of cases exhibit patterns of evolution and extension that would have
been wholly unpredictable at the decision of the first case or even at its first, second, or third
application.
121See Brest, supra note iS, at 231-34.
122See id. at 222-24, 231-34.
123This characterization of the ambition of moderate interpretivism is not the only one
possible. There are at least two alternatives, both exhibited by the work of Robert Bork. In
an important article written in 1971, Bork argued that constitutional principles, in order to be
legitimate, must be "neutral" not only in their substance but also in their derivation. See Bork,
supra note 52, at 7. The implication was that value arguments, as I have defined them, should
be excluded from the constitutional calculus. See id. at 7-8. In subsequent writing, however,
Bork has sometimes retreated from this position. For example, in a recent speech Bork argued
that moderate interpretivism recognizes the need for judges to appeal to arguments of value in
giving content to vague constitutional principles such as those embodied in the first amendment's
1216
(Vol. 1oo:1189
19871
CONSTRUCTIVIST COHERENCE
1217
that no state may deny any person the equal protection of the laws
a principle susceptible to applications far more expansive than
those that the framers themselves would specifically have endorsed. 128
Once a consistent adherence to the framers' specific intent is renounced, there simply is no value-neutral way to choose among pos12 9
sible specifications of the framers' abstract intent.
Recognition that moderate interpretivism relies on factors besides
text and history - and in particular that judges must make delicate
moral and political judgments to determine the contemporary meaning
of numerous constitutional concepts and the level of abstraction at
which the framers' intent ought to be stated - reveals the theory as
failed in its own terms. 130 Arguments from text and the framers'
intent turn out not to perform an independent and privileged function
of restraining judicial choice and determining constitutional outcomes.
Other factors not only do, but must, figure crucially in the constitutional calculus. 131
2. Ely's Democracy and Distrust Theory. No recent constitutional theory has generated more interest and enthusiasm than that
offered in Democracy and Distrust by Dean Ely. 13 2 Ely's theory
shares much with interpretivism.1 33 In common with interpretivists,
128 See R. DWORKIN, TAKING RIGHTS SERIOUSLY, supra note 37, at 133-36. For example,
this formulation could support changes in the legal treatment of women or of homosexuals.
129 Judge Bork has defended Brown as being grounded in the "neutral" principle that racial
discriminations were unconstitutional. See Bork, supra note 52, at 14-15. But Judge Bork's
selection of the principle to be enforced - founded on the idea that the core purpose of the
framers was to establish racial equality - itself reflects a contestable choice among the levels
of abstraction with which their purposes might be stated. See Brest, The FundamentalRights
Controversy: The Essential Contradictionsof Normative ConstitutionalScholarship, 90 YALE
L.J. io63, 1090-91 (198).
130 See supra note 123.
131 It does not follow, however, that nothing remains of the rule of law. It is a mistake to
believe, as the interpretivist does, that "the rule of law" requires judges' decisions to be "determined" by either the constitutional text or by the framers' intent, or even to be independent of
the judge's moral and political beliefs. The failure of interpretivism does not mean that there
is no rule of law. It reveals instead that the rule of law does not mean what the interpretivist
thought it did and that its meaning must be sought, not in stipulative definitions, but in the
social and legal contexts in which the phrase has its most natural use. See generally L.
WITTGENSTEIN, PHILOSOPHICAL INVESTIGATIONS (G. Anscombe trans. 3d ed. 1976). I have
no pretension of being able to develop here a theory of the rule of law - undoubtedly a task
of enormous difficulty. Yet clearly the starting point should not be the interpretivist's stipulative
definition. For an insightful probing of some of the leading issues, see Michelman, cited in note
99 above.
132See generally Symposium: Constitutional Adjudication and Democratic Theory, 56
N.Y.U. L. REv. 259 (I98i); Symposium: JudicialReview versus Democracy, 42 OHIo ST. L.J.
I (1981).
133 Indeed, Ely recognizes that his theory could be viewed as a species of interpretivism.
See J. ELY, supra note 6, at 12-13, 87-88. But see Bork, Styles in ConstitutionalTheory, S.
1218
[Vol. IOO:ii8 9
383, 390 (1985) (arguing that Ely "is a non-interpretivist, whether he knows it or
he tries to make "the Constitution more democratic than the Constitution is").
ELY, supra note 6, at 4, 41.
at 41.
at 3.
at 13, 30. In this regard, the various constitutional provisions can be arrayed
along a spectrum.
Some, like the requirement that the President be 35 years old, are highly
determinate. Others, like the first and eighth amendments, provide at least a structured starting
point for constitutional reasoning yet "cannot intelligibly be given content solely on the basis of
their language and surrounding legislative history." Id. at 12. Still others are totally openended and allow much imposition of values from outside of the Constitution.
138 J. ELY, supra note 6, at 4I.
139 For an illuminating discussion of some alternative readings, see Estreicher, Review Essay,
Platonic Guardians of Democracy: John Hart Ely's Role for the Supreme Court in the Constitution's Open Texture, 56 N.Y.U. L. REv. 547 (1981).
140 J. ELY, supra note 6, at 34.
141 Id. at 32.
142 Id.
143 In Ely's view, it is so offensive for judges to choose without constraint among value
arguments to give meaning to the open-ended guarantees that, if there were no other alternative,
"responsible commentators must consider seriously the possibility that courts simply should"
refuse to enforce the open-ended provisions altogether. Id. at 41.
144 See id. at 87-88, 102-04, 181-83.
19s71
CONSTRUCTIVIST COHERENCE
121
1220
[Vol.
IOO:I I89
I6,
17,
28,
to dismiss
'the intent of the framers' as beside any relevant point .... [S]ometimes in order to know what
was ratified we need to know what was intended.").
152 See id. at 70.
153 For forceful criticisms of Ely's theory, see Parker, The Past of Constitutional Theory And its Future, 42 OHIO ST. L.J.
223
ConstitutionalTheories, 89 YALE L.J. 1o63, IO64 (Ig8o); Tushnet, Darkness on the Edge of
Town: The Contributions of John Hart Ely to ConstitutionalTheory, 89 YALE L.J. 1037 (1980).
154 See Parker, supra note 153, at 221 n.4; Tribe, supra note 153, at io67-77.
155 See J. ELY, supra note 6, at 221 n.4.
156 Id. at 41.
157 See id. at 88-oI.
M55
See Parker, supra note 153, at 236 ("Process itself, therefore, becomes substantive.")
(citing Tribe, supra note 153, at 1070-72).
159 See L. TRIBE, supra note 96, at io-ii; Schauer, supra note 17, at 816-i9.
160 See L. TRIBE, supra note q6, at io-ii; Schauer, supra note 17, at 816-17.
1987]
CONSTRUCTIVIST COHERENCE
1221
to prefer it. In short, support for the theory must come from sources
external to the text - from arguments of value very like the kind
that Ely purports to reject as illegitimate. 161
Beyond their necessary role in the derivation of Ely's theory, value
arguments have a similar and equally problematic part in its application.1 62 An example will illustrate that "substantive" and "process"
values simply cannot be kept apart and that arguments about substantive values inevitably creep into the constitutional calculus. Ely
argues that courts should invalidate substantive legislation under the
equal protection clause only to correct process failures that result from
prejudice against minorities. 163 His theory seems to raise no objection
to legislation that enforces moral judgments that minorities find repugnant; he contends only that minorities must have a fair opportunity
to participate in the process by which the legislation is produced.164
The difficulty is that Ely's theory gives no basis for distinguishing
moral judgments, which will support a legislative choice, from prej5
udice, which will not. 16
The difficulty of the distinction is illustrated by a case that Ely
himself discusses: legislation penalizing homosexual conduct.1 66 How
are we to tell whether such legislation reflects a permissible moral
judgment that homosexual acts are wrong or results instead from
impermissible prejudice? Ely suggests that "prejudice" may be de-
161See Parker, supra note 153, at 235 ("Ely's theory, then, appears to me to be Ely's theory
- no more and no less. As such, it is very interesting. But by Ely's own standard that is not
enough."). Ely, of course, is too sophisticated not to recognize this. See J. ELY, supra note 6,
at IoI. Indeed, beyond the claim that his theory describes the Constitution accurately, Ely
relies on what he labels "overtly normative" arguments. Id. But while Ely claims that his
normative arguments are different from those of the judges and theorists whose substantive
value judgments he objects to, his effort at distinction is unpersuasive.
First, Ely defines judicial "value-imposition," id. at 75 n.*, the evil he seeks to avoid, in
such a way that his preferred "process" or "participational" values fall outside the scope of the
term. See id. It is one thing, he says, to value certain substantive outcomes and quite another
to endorse judicial enforcement of process or participational values. See id. This, however, is
less an argument than a definitional fiat, and Ely himself seems to recognize as much. See id.
Not everyone values democracy as much as Ely does. Moreover, "democracy" is an essentially
contestable concept, which cannot provide a basis for judicial action without an infusion of
normatively controversial content.
Second, Ely argues that his preferred process values - in contrast with the values relied on
by others - are those that "the Court should pursue." Id. But this purely normative argument,
whatever its independent force, compromises rather than supports the attempted distinction
between arguments asserting process values and those concerned with substantive values: the
suggestion that the substantive and participational values can be weighed against each other
according to a uniform scale implies that there is no difference in kind.
162 See L. TRIBE, supra note 96, at I1-19.
163 See J. ELY, supra note 6, at 153.
164 See id.
165 See L. TRIBE, supra note 96, at I1-19.
166 See J. ELY, supra note 6, at 162-64.
1222
1987]
CONSTRUCTIVIST COHERENCE
1223
1224
[Vol. IO0:II8 9
participants in any complex practice will necessarily outrun any rule-like statement of that
knowledge, see M. POLANYI, PERSONAL KNOWLEDGE 69-131 (1958).
175 Rules and principles are usefully distinguished. See R. DWORKIN, TAKING RIGHTS
SERIOUSLY, supra note 37, at 22-28. A rule, if it applies to a situation, must be obeyed. By
contrast, principles can apply to a situation yet be outweighed by other, competing principles.
For a similar distinction between rules and "standards," see Kennedy, Form and Substance in
Common Law Adjudication, 89 HARv. L. REV. 1685, 1687-1701 (1976).
176 See Fish, Fish v. Fiss, supra note 26, at 1333-34.
177 See id. at 1328-32.
178 See L. TRIBE, supra note 96, at 3-1I.
179 See, e.g., Fish, Fish v. Fiss, supra note 26; Fish, Chang Gang, supra note 26.
180 Open-system theories need not be wholly nihilistic. It is quite consistent with an opensystem approach to recognize the possibility of useful generalizations about what ordinarily is
done in our practice of constitutional interpretation. Open-system theories, as I have defined
them, need only deny the possibility of accurate rule-like statements specifying what someone
must do in order to conform to the norms of the practice of constitutional interpretation. Nor
do open-system theories deny that constitutional interpreters experience a sense of constraint
arising from the nature of the practice in which they are engaged. They insist only that the
constraint is better characterized in terms of a "tacit awareness of what is possible," Fish, Chain
Gang, supra note 26, at 562, than in terms of general rules or norms.
19871
CONSTRUCTIVIST COHERENCE
1225
181Open-system theories need not deny that constitutional decisionmakers normally accept
guidance from such "informal criteria" as habits, nonbinding conventions of practice, intuitions
that some outcomes will be perceived as normatively more attractive than others, and commonly
shared understandings about proper institutional roles. See generally Stick, Can Nihilism Be
Pragmatic?, OO HARv. L. REV. 332, 354-58 (1986) (discussing possible bases for predictability
in law). Nor do they deny the possibility of predictability in our constitutional practice. They
insist only that predictions about how different kinds of arguments will be combined or weighed
against each other in particular cases could not be based on the assumption that judges and
lawyers will conform to identifiable rules or principles. Instead, predictions would need to be
based on other factors. See id.
182 See, e.g., L. TRIBE, supra note 96, at 3-11; Fish, Fish v. Fiss, supra note 26; Fish,
1226
[Vol. ioo:ii89
186 Professor Fish appears to believe the contrary: that even if lawyers, judges, and justices
recognized certain rules of interpretive practice, the rules would be virtually meaningless. The
rules could not "constrain" argument or decisionmaking, he argues, because the rules themselves
would always have to be "interpreted" by the very persons they were intended to constrain.
See Fish, Fish v. Fiss, supra note 26, at 1326-27. But this argument fails to establish either
that the practice of constitutional interpretation lacks rules or that its rules are insignificant.
For rules to be effective, a constitutional interpreter need only be capable of following them in
roughly the same way that rules are followed in other contexts and practices. Sophisticated
hermeneutic theories suggest that rule-following may be a more complicated and even a more
precarious process than we might have thought, but not that we cannot sensibly refer to certain
conduct - including that of judges and even justices of the Supreme Court - as rule-following.
See generally L. WITTGENSTEIN, supra note 131.
19871
CONSTRUCTIVIST COHERENCE
1227
1228
[Vol. lOO:1189
1987]
CONSTRUCTIVIST COHERENCE
1229
and against upholding a claim of constitutional right. See, e.g., Gunther, In Search of Judicial
Quality on a Changing Court: The Case of Justice Powell, 24 STAN. L. REV. 1001 (1972)
(discussing a balancing approach as employed by Justice Powell); Mendelson, On the Meaning
of the First Amendment: Absolutes in the Balance, 50 CALIF. L. REv. 821 (1962) (discussing a
balancing methodology used in first amendment cases). There are also conspicuous opinions in
which courts have engaged in conspicuous balancing. See, e.g., Mathews v. Eldridge, 424 U.S.
319 (1976) (fifth amendment); Konigsberg v. State Bar, 366 U.S. 36 (i96i) (first amendment).
But the type of balancing that is commonly discussed and practiced is balancing within a
category of argument. First amendment balancing theories, for example, do not generally
suppose that arguments from the text point one way, arguments of the framers' intent another,
and so forth, and then prescribe balancing of one categorical conclusion against another to
resolve the impasse. Instead, the familiar balancing theories operate within the category of
arguments of value or of constitutional theory, where they attempt to identify and weigh
competing interests. See, e.g., Gunther, supra.
192 A theory like balancing seems a strong candidate to reflect the preferred approach of
"non-interpretivists" or "supplementers," see Grey, The Constitution as Scripture, 37 STAN. L.
REv. 1, 1-5 (1984), who argue that the courts should take account of arguments from text and
the framers' intent but should not deem them the controlling authority in every case. Unfortunately, proponents of this approach have seldom taken the next step of demonstrating how
arguments from text and the framers' intent ought to be weighed against other kinds of arguments to arrive at an outcome in a particular case.
1230
[Vol. 1oo:1189
93
framers' intent. 1
1987]
CONSTRUCTIVIST COHERENCE
1231
Together, the various types of criticism will help to suggest the criteria
that a better theory needs to satisfy.
One important type of criticism involves descriptive adequacy.
Privileged factor theories deny the relevance of all but one or two
factors. Yet at least five different types of argument are too widely
recognized as legitimate for a privileged factor theory to be a plausible
theory of our existing constitutional practice. A role for all five categories of argument must be recognized. But those roles are not the
ones indicated by a balancing approach. Balancing also misdescribes
what actually happens in constitutional interpretation. Rarely if ever
do judges and lawyers indicate that different factors point to different
conclusions or that the factors must be balanced against each other
to determine the correct result.
A second line of criticism has been expressly normative. My normative criticisms have come from two sides. On the one hand, I have
criticized originalism, which rejects any attempt to accommodate modern needs and evolving moral and political sensibilities by departing
from the framers' specific intent, for being too rigid and inflexible.
On the other, open-system theories have nearly the opposite vice.
They provide insufficient" structure to constitutional argument and
interpretation. Some middle position is needed.
Finally, a third critique has been that the various kinds of constitutional argument are not as independent as several approaches to the
problem of commensuration assume. Privileged factor and balancing
theories are both vulnerable to this criticism. Value arguments, in
particular, permeate the other categories of argument. But there are
other interconnections as well. A more successful approach to the
commensurability problem would need not only to recognize interrelationships and interdependences among the various categories of argument but also to trace the lines along which arguments concerning
one factor can influence judgments concerning another.
[Vol. ioo:ii89
200 For philosophical analyses of the concept of a social practice, see A. MACINTYRE, AFTER
VIRTUE 175-89 (i981); Rawls, Two Concepts of Rules, 64 PHIL. REV. 3 (1955); Taylor, Inter-
pretation and the Sciences of Man, in INTERPRETIVE SOCIAL SCIENCE: A READER 25, 46-55
(P. Rabinow & W. Sullivan eds. 1979).
201 Interchangeable use of the terms sacrifices the precision that might be gained by rigorous
adherence to a distinction between "rules," which must be observed if they apply at all, and
"standards" (and possibly "norms"), which can be outweighed by other articulable factors also
relevant to a decision. See supra note 175. In failing to differentiate the terms here, I mean
to emphasize what they have in common: "rules," "norms," "standards," and "conventions" all
are publicly statable guides to conduct that are capable of giving structure to a practice and of
supporting judgments that conduct and arguments are intelligible or unintelligible, appropriate
or inappropriate, good or bad, or right or wrong.
202 See J. SEARLE, SPEECH ACTS 42 (1969).
1987]
CONSTRUCTIVIST COHERENCE
1233
that the body politic can at a specific point in time definitively order relationships, and
that such an ordering is binding on all organs of government until changed by amendment")
with R. DWORKIN, supra note II, at 359-69 (rejecting "historicism" as a theory of constitutional
interpretation).
...
theories on the ground that conventions will run out before giving needed guidance in hard
cases); Fish, Fish v. Fiss, supra note 26, at 1330-31; Levinson, supra note 26, at 401-03.
20s See supra pp. 1213, 1222-23, 1225-26, 1229-30.
206 Although I follow Dworkin in seeking an interpretive theory of our constitutional practice,
I believe that I diverge from his approach in offering a theory that relies on rules or conventions
1234
1987]
CONSTRUCTIVIST COHERENCE
1235
supra note 207. Dworkin argues that because law is an "interpretive concept," lawyers must
develop their own individual theories of the practice in which they are engaged in order to
know what constraints and obligations the practice imposes. See R. DWORKIN, supra note i i,
at 45-86. In developing such theories, Dworkin argues, the principal task is to develop an
account of the purposes of legal practice. See id. at 50-52. With the interpretive task defined
to include the imposition of purpose on the practice of legal interpretation, normative commitments become crucial to legal theory in two ways. First, an interpreter's moral and political
views will influence her understanding of the practice in which she is engaged, because it is
part of the essence of interpretation to characterize the interpreted practice as "the best possible
example of the form or genre to which it is taken to belong." Id. at 52. Second, where the
criteria of "fit" and normative excellence are in tension, the interpreter's moral commitments
will influence her judgment of where the balance ought to be struck. See id. at 52-53; R.
DWORKIN, A IATTER OF PRINCIPLE, supra note 26, at 16o-62.
214 See supra pp. 195-98 (discussing arguments from text).
21 See inf a p. 1238 (discussing the influence of precedent on arguments from text).
1236
[Vol. ioo:ii8 9
at 45-113.
218 See Simon, The Authority of the Constitution and Its Meaning: A Preface to a Theory
of ConstitutionalInterpretation,58 S. CAL. L. REv. 603, 613-14 (i985).
219 See Rawls, supra note 2oo, at 27-28. This point needs qualification. Although I have
suggested that criteria of evaluation emerge from societal values and needs, it is a mistake to
assume either that societal values can be identified other than through an interpretive process
1237
CONSTRUCTIVIST COHERENCE
1987]
In this Part I offer a solution to the problem of the commensurability of different kinds of constitutional argument. That solution
emerges from an interpretive theory of our constitutional practice that
I call a constructivist coherence theory. Although this Part provides
a full introductory statement of constructivist coherence theory, the
discussion is frequently abstract and summary. Subsequent Parts will
further demonstrate the validity of the empirical assumptions on which
constructivist coherence theory depends, more fully establish its workability, and show its normative and descriptive power. The burden
of this Part is to frame the theory as a whole and thereby to establish
the structure against which more detailed argumentation can occur in
Parts V and VI.
or that society can be viewed as wholly independent of law in order to determine what form of
constitutional practice will best serve society. The reason is that our constitutional law and
interpretive practice are partly constitutive of society, just as, on the other hand, they are partly
Change in our constitutional practice will change society, leaving it
constituted by society.
REV. 1276, I288-89 (1984). At the same time, however, the artificial division of law and society
can be justified by its usefulness in helping us to reason about and improve our law and legal
practice.
220 See S. CAVELL, MUST WE MEAN WHAT WE SAY? 24-30 (1969); Rawls, supra note 200,
at 4, 26-27, 29-30, 32.
1238
[Vol. lOO:1189
A. Empirical Foundations
In developing a theory of how the different kinds of constitutional
argument fit together in our practice of constitutional interpretation,
I begin with three reference points, all of which possess relatively firm
empirical foundations. First, all of the five kinds of argument discussed in Part II have an accepted place in constitutional argument.
Any theory that denied the legitimacy of one or more of them would
mischaracterize our constitutional practice.
Second, although the categories of argument are sufficiently independent to merit separate exposition, there are numerous interconnections among them. Perhaps most importantly, arguments of value
infuse arguments and influence conclusions within other categories.
For example, arguments from text at least sometimes involve essentially contested concepts, the meanings of which cannot be specified
independently of moral and political assumptions 22 1 and, if a challenge
is mounted, of moral and political argument. More commonly, arguments within the categories of the framers' intent, constitutional
theory, and precedent presuppose a theory of the meaning of the
pertinent legal materials that will necessarily reflect value judgments
222
or value-influenced choices.
Other conceptual interrelationships among the categories are also
important. For example, arguments from text cannot be held rigidly
separate from arguments based on precedent; the range of plausible
and permissible legal constructions of a word, phrase, or provision is
affected, at least at the margins, by prior judicial decisions interpreting
the relevant language. An example comes from due process jurisprudence. Although a linguistically plausible argument from text holds
that "'substantive due process' is a contradiction in terms - sort of
like 'green pastel redness,' '223 a number of respected Supreme Court
decisions reject this position. 224 Although these cases may rely on
implicit or explicit assumptions about the framers' intent, the existence
of the precedents themselves is independently relevant to the question
of whether "substantive due process" methodology is appropriate in
future cases. 225 Another instance of interrelationship among the categories involves the dependence of arguments of constitutional theory
on constructions of constitutional text. Because constitutional theory
221
222
223
224
225
increasingly persuasive evidence of the framers' contrary intent before that factor will be allowed
to prevail against other kinds of argument. For a discussion of the relative authority of the
framers' intent in comparison with other kinds of arguments, see pages 1243-46, 1254-58 below.
1987]
CONSTRUCTIVIST COHERENCE
1239
presupposes an interpretation of textual language, and because opentextured language invites construction to accord with theory, blurring
occurs in some cases between the two categories.
In addition to the necessary connections between the various categories of constitutional argument, the potential for further interdependence arises from the incommensurability of competing arguments
within certain of the categories. Incommensurability obtains when
there are competing conceptions of what the arguments within a
particular category properly refer to. 2 2 6 One example emerges from
the category of arguments from text and the competition between the
original understanding and contemporary meaning. Another, within
the category of the framers' intent, involves the opposing claims of
specific and of abstract intent. As I have argued before, no standard
that is purely internal to those categories can be applied to choose,
on a case-by-case basis, between arguments that depend on competing
conceptions of what the categories are or ought to be about. 2 27 The
requisite choices could be made, however, by reference to other categories of argument. It would be a plausible selection principle to
prefer, in any particular case, the conception of arguments from text
or the framers' intent that would cohere best with the result prescribed
by arguments in other categories. A similar criterion of choice could
be used when the arguments within a category, though commensurable, are indeterminate - when they are in near equipoise or otherwise susceptible of rethinking and adjustment.
My third empirical reference point involves the striking extent to
which judges, lawyers, and other constitutional interpreters find the
various factors of constitutional argument to accord, or at least to be
consistent with their prescription of, a single result. 2 28 Only very
rarely do legal arguments - including those of judicial, and in particular Supreme Court opinions - find the best arguments within
different categories to point in divergent directions. It is possible, of
course, that the coherence exhibited by legal briefs, judicial opinions,
and scholarly articles evidences nothing more than an argumentative
226"Incommensurability," which implies that claims made against one background of assumptions cannot be translated without loss into the argumentative framework structured by a
different set of assumptions and beliefs, is not the same thing as "incomparability." See R.
BERNSTEIN, BEYOND OBJECTIVISM AND RELATIVISM: SCIENCE, HERMENEUTICS, AND PRAXIS
82 (1983); Cornell, Taking Hegel Seriously: Reflections on Beyond Objectivism and Relativism,
7 CARDOZO L. REv. 139, 152-6o (1985). Even though the claims about the original understanding may be incommensurable with claims about current meaning, the two conceptions can be
usefully and rationally compared and a judgment made to prefer one or the other. Indeed,
constructivist coherence theory is in part a theory of how such competing conceptions within
the categories of constitutional argument are and ought to be compared for purposes of choice
in a particular case.
227See supra p. 1215.
226 See supra p. 1193.
1240
[Vol. 1oo:1189
tactic that conceals the advocate's deep understanding of our constitutional practice. This is an issue to which I shall return. 229 For
now, however, the almost total lack of dissonance generally acknowledged among the various types of constitutional argument is all that
I mean to insist upon.
B. The Coherence Thesis
The striking degree of coherence of arguments exhibited in the
typical constitutional argument, in conjunction with the interconnectedness of the various relevant factors, suggests the principal hypothesis
that constructivist coherence theory embodies. The implicit norms of
our practice of constitutional interpretation prescribe an effort to
achieve plausible understandings of arguments from text, the framers'
intent, constitutional theory, precedent, and relevant values, all of
which point to the same result. A provisional conclusion may be
reached as to the balance of argument within each factor; perhaps
more commonly, the decisionmaking process will have a gestalt-like
quality, in which each category is considered with all of the others in
mind. But if the conclusions fail to cohere into a uniform prescription
for how the case or issue ought to be resolved, then any or all of the
individual conclusions may be reexamined, and the results adjusted
insofar as plausible within the prevailing conventions of constitutional
230
analysis, in an effort to achieve a uniform outcome.
229 See infra pp. 1248-49.
calls for a two-way traffic of adjustment and correction between intuitive judgments about what
is right in particular cases, and the moral principles that define what is right in general. When
a general principle and a moral intuition collide, some correction is needed in either the principle
or the particularistic judgment. Yet one cannot specify a priori in which the adjustment ought
to occur. All that the methodology prescribes is that coherence should be sought either by
reformulating the governing principle or by revising the judgment about what is right in the
particular case. See R. DWORKIN, TAKING RIGHTS SERIOUSLY, supra note 37, at 159-68;
Feinberg, supra, at 1oxS-21. John Rawls's concept of "reflective equilibrium," J. RAWLS, A
THEORY OF JUSTICE 20-22, 48-53 (1971), calls for a process of adjustment and correction to
achieve coherence that establishes his as a coherence theory. See R. DWORKIN, TAKING RIGHTS
SERIOUSLY, supra note 37, at 159-68; Feinberg, supra, at ioi8-2i. But the coherence approach
has intellectual antecedents running back to Aristotle. See S. HAMPSHIRE, TWO THEORIES OF
MORALITY 6-28 (1977).
In common with coherence theories of moral and political philosophy, constructivist coherence
1987]
CONSTRUCTIVIST COHERENCE
1241
theory assumes a decision calculus that assesses the rightness or wrongness of possible conclusions
along more than one dimension. As a further point of analogy, constructivist coherence theory
assumes that the various relevant factors of argument are internally related in a way that
frequently makes it possible, through adjustment and accommodation, to achieve coherence or
accord. But constructivist coherence theory is a theory of legal practice and decisionmaking,
not of pure moral and political decisionmaking. As a result, constructivist coherence theory
departs sharply from moral and political coherence theories in its characterization of the kinds
of factors among which coherence should be sought.
Constructivist coherence theory is also "constructivist" in a dual sense. First, it assumes that
conclusions about what are the best arguments reflect standards that are internal to the practice
of constitutional interpretation - and in that sense are human constructs - rather than
standards that are dictated by some external and objective reality to which legal argument and
conclusions should strive to correspond. There is thus some similarity between its assumptions
and those of philosophical theories, frequently described as "constructivist," that characterize
our entire understanding of the social and even the physical universes as being socially constructed rather than as either mirroring or being determined by nature. See, e.g., T. KUHN,
THE STRUCTURE OF SCIENTIFIC REVOLUTIONS (2d ed. 197o); R. RORTY, PHILOSOPHY AND THE
MIRROR OF NATURE (1979); THE INVENTED REALITY (P. Watzlawick ed. 1984). Second, the
coherence achieved in any particular case may be more or less purposively "constructed" in a
situation in which dissonant conclusions concerning what are the best arguments within particular categories might initially have appeared natural or correct.
231 See Katz v. United States, 389 U.S. 347, 35o-53 (1967).
232 See supra p. 1238.
2-3 See, e.g., Home Bldg. & Loan Ass'n v. Blasdell, 290 U.S. 398 (1934); Epstein, Toward
1242
[Vol. 1oo:118
19871
CONSTRUCTIVIST COHERENCE
1243
1244
[Vol. lOO:1189
236 Cf. McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 407 (1819) (Marshall, C.J.) (stating
that "we must never forget, that it is a constitution we are expounding" (emphasis in original)).
237 See Grey, Origins of the Unwritten Constitution: FundamentalLaw in American Revolutionary Thought, 30 STAN. L. REv. 843, 844 (1978).
238 Arguments from text may, of course, exclude some meanings without conclusively resolving which among the plausible meanings ought to be preferred.
239 There are some situations in which an argument from text might in fact be dispositive.
See, e.g., infra pp. 1283-85 (discussing the constitutionality of slavery under the Constitution
several examples. See, e.g., Williams v. Florida, 399 U.S. 78, 92-100 (1970) (asserting that
"history" shows the jury trial provision was to be understood functionally and therefore twelve
jurors, the historical number, are not required); Bell v. Maryland, 378 U.S. 226, 288-312 (1964)
(Goldberg, J., concurring) (arguing in purported reliance on original intent that the fourteenth
amendment prohibits privately owned restaurants from discriminating on grounds of race); cf.
Brest, supra note 18, at 234-35 ("Although the Supreme Court has often disavowed strict
originalism and acknowledged its moderate originalist stance, it has not usually admitted that
some of its decisions go still further." (footnotes omitted)).
1987]
CONSTRUCTIVIST COHERENCE
1245
242
4. Arguments from precedent. - Although constitutional arguments sometimes turn almost completely on disputed interpretations
of judicial precedents, the principle is clear that stare decisis is entitled
to less weight in constitutional than in nonconstitutional cases, 24 3 and
the Supreme Court stands ready to "correct its errors even though of
long standing. '244 Precedent therefore occupies the fourth place in the
hierarchical order. This status could easily prove misleading, however, and thus requires two short comments. First, because arguments
in the categories ranked above precedent often prove indeterminate
in hard cases, it would be a mistake to think the significance of
precedent will be only marginal. On the contrary, its ranking is
wholly consistent with the observation that as the text "gets older and
interpretive materials accumulate, the focus of professional and judicial opinion shifts from .
1246
[Vol. lOO:1189
19871
CONSTRUCTIVIST COHERENCE
1247
247 See
1248
[Vol. IO0:Ii8 9
1987]
CONSTRUCTIVIST COHERENCE
1249
249 See Grey, supra note 18, at 706 ("[I]f judges resort to bad interpretation in reference to
honest exposition of deeply held but unwritten ideals, it must be because they perceive the latter
mode of decisionmaking to be of suspect legitimacy.").
1250
[Vol. lOO:1189
250 But
19871
CONSTRUCTIVIST COHERENCE
1251
1252
[Vol. ioo:ii8 9
This Part undertakes to show in greater detail how the five categories of constitutional argument lend themselves to the kinds of
accommodation prescribed by constructivist coherence theory. To accomplish this end, the categories are discussed in turn. With respect
to each, I argue that the initial balance of arguments within the
category - if the categories were conceived as sharply independent
from one another - would be likely either to raise an incommensurability problem that could not be resolved without appeal to some
outside arbiter or to prove substantially indeterminate or uncertain.
With incommensurability and indeterminacy limiting the autonomy of
the individual categories, the gestalt-like quest for coherence that is
modeled by constructivist coherence theory not only becomes possible
but fulfills an important function in our constitutional practice. Although the main thread of argument addresses the workability and
the descriptive power of constructivist coherence theory, I also argue
that constructivist coherence theory characterizes our actual practice
of constitutional interpretation in the best plausible light within the
structure of social values and expectations in which our constitutional
practice is situated.
A. Arguments From Text
Three characteristics establish arguments from text as well accommodated to constructivist coherence analysis. The first involves the
rivalry within the category between the claims to authoritativeness of
the original understanding on the one hand and contemporary meaning on the other. In the context of our current practice, it is simply
impossible to maintain that original historical meanings are uniquely
relevant. Choices between
original and contemporary understandings
25 2
must therefore be made.
252 The phenomenon of change in meanings is easily documented.
note 18, at 710-14; Munzer & Nickel, supra note 30, at 1029 n.Io. It is among the virtues of
1987]
CONSTRUCTIVIST COHERENCE
1253
A second feature of arguments from text that suits them to constructivist coherence theory arises from the relationship between "ordinary language" on the one hand and "constitutional language" on
the other. Although constitutional discourse depends heavily on the
understandings implicit in our use of ordinary language, the meanings
of certain constitutional terms diverge significantly from the meanings
of those same terms in ordinary speech. 253 Leading examples come
from substantive due process 254 and fourth amendment jurisprudence. 255 As a result, the possibility that a constitutional term might
be either a term of art or an instance of more ordinary usage some-
arguments for or against constituting the practice of constitutional interpretation in such a way
that meanings are susceptible to alteration, it is mistaken to urge either the descriptive claim
that the Constitution's meaning does not, or the logical claim that it cannot, alter.
253 See Schauer, supra note 17, at 799-801. This is but an instance of the more general
insight that meaning depends on context. For classic statements, see generally J. AUSTIN, How
TO Do THINGS WITH WORDS (J. Urmson & M. Sbisa 2d ed. 1975); J. AUSTIN, PHILOSOPHICAL
PAPERS (J. Urmson & E. Warnock 3d ed. 1979); L. WITTGENSTEIN, supra note 131.
254 As the Supreme Court recently recognized:
[D]espite the language of the Due Process Clauses of the Fifth and Fourteenth Amendments, which appears to focus only on the processes by which life, liberty, or property
is taken, the cases are legion in which those [cilauses have been interpreted to have
substantive content, subsuming rights that to a great extent are immune from federal or
state regulation ....
Bowers v. Hardwick, io6 S. Ct. 2841, 2844 (1986). Controversy persists less over the general
point than over the question of what rights are so protected.
255See, e.g., Katz v. United States, 389 U.S. 347, 350-53 (1967) (holding electronic eavesdropping to be a "search and seizure" within the meaning of the fourth amendment).
256 See Moore, supra note 15, at 313-38; cf. Hardwick, io6 S. Ct. at 2846 ("The Court is
most vulnerable and comes nearest to illegitimacy when it deals with judge-made constitutional
law having little or no cognizable roots in the language or design of the Constitution.").
257 Cf. A. Cox, THE ROLE OF THE SUPREME COURT IN AMERICAN GOVERNMENT 113 (1976)
(despite the linguistic oddness of finding substantive rights protected by the due process clause,
the Supreme Court's "persistent" invocation of substantive due process notions "for almost a
century attests the strength" of the natural law impulse supporting the concept, such that it
would be "unwise as well as hopeless to resist it").
25SThe idea of an "open texture!' is developed in H. HART, THE CONCEPT OF LAW 121-32
(196i).
1254
[Vol. IOO:xI89
essentially contestable character of much constitutional language. Although even highly general constitutional provisions typically exclude
some interpretations, 25 9 constitutional language often will support or
bear more than one reading. 260 In such cases, nontextual factors may
guide selection among the plausible alternatives.
The normative support for this interpretive account of the role of
arguments from text is largely implicit in what has gone before.
Where arguments from text are indeterminate or tentatively point
toward normatively questionable results, it is desirable to give open
weight to factors with independent claims to recognition. At the same
time, however, by recognizing that arguments from text head the
hierarchy of relevant factors, constructivist coherence theory promotes
a conception of constitutional law as something distinct from, even if
properly influenced by, judicial visions of sound policy and moral
justice.
B. Arguments of Historical Intent
Two features of the concept of the framers' intent make it amenable to a constructivist coherence analysis. One involves the indeterminacy inherent in efforts to ascertain the framers' specific intent.
The other stems from the distinction between specific intent and abstract or general intent.
Although a complete defense of this claim would take me far
beyond the sensible limits of this paper, I have argued that the concept
of the framers' specific intent can function helpfully in constitutional
interpretation only if it is viewed in each case as a theoretical construct, developed through a process of interpretation, that expresses
the interpreter's conclusion concerning what set of moral and political
views best explains the adoption of a constitutional provision in light
of all the available historical evidence. 26 1 Thus defined, the notion of
specific intent lends itself to constructivist coherence analysis, because
the historical evidence frequently will be so uncertain or conflicted as
to be consistent with several interpretations. 262 This is especially
likely to be true where the framers chose vague or open-ended language or where their intent must be projected onto a situation that
the framers themselves never foresaw. Under such conditions, other
factors may either establish a preference among the contending historical conclusions or, in the alternative, provide ample reason for
259 See Schauer, supra note 17.
260 Cf. White, supra note 26, at 415 (asserting that reading a legal text is often not so much
reading for a single meaning as reading for a range of possible meanings).
261 See supra pp. 1212-13.
262 For discussion of a concrete example, see infra text accompanying notes 331-33.
1987]
1255
264 See supra pp. 1216-17 (discussing the problem of choice among the different levels of
abstraction at which intent might be identified).
265 See supra p. 1244.
266 Indeed, I am aware of only one case, Reynolds v. Sims, 377 U.S. 533 (1964), in which
the Court failed to reject the contention of a dissenting opinion that its conclusion could not be
reconciled with the intent of the framers. But the exception only proves the rule: defenders of
the Reynolds result have generally agreed with critics that the dissenting opinion's historical
arguments required an answer. See, e.g., Van Alstyne, The Fourteenth Amendment, the "Right"
to Vote, and the Understandingof the Thirty-Ninth Congress, 1965 SuP. CT. REv. 33, 34-38.
A much more common gambit is for the Court to assert or imply that the framers, whatever
their specific intentions, would have intended the courts to adapt constitutional principles to
changing situations. See, e.g., Home Bldg. & Loan Ass'n v. Blaisdell, 290 U.S. 398, 443 (1934)
(suggesting that the framers of the contract clause, as of all "the great clauses of the Constitution,"
would have contemplated adaptation to various crises, and finding "no warrant for the conclusion
that .. . the founders of our Government would have interpreted the clause differently [than
the Court had done] had they had occasion to assume that responsibility in the conditions of
the later day").
1256
[Vol. IOO:II89
In seeking an attractive description, however, there are good reasons to retain a role for the framers' intent. Our Constitution is not
and should not be conceived as loosed conceptually from the purposes
and intentions of those who drafted and ratified it. The claim of the
framers' intent to argumentative authority is especially strong with
recent constitutional amendments. When an amendment is newly
enacted, arguments that the value of democracy requires respect for
the intent of the framers and ratifiers are especially strong. 26 7 Even
with older provisions, however, the arguments for respecting the framers' intent never dissipate entirely. Continued acknowledgment of
this factor's authority is based on a recognition that our living Constitution is the product of a tradition in which the present cannot be
understood independently of the past. Indeed, our constitutional practice presupposes that it is only on the basis of a text produced by past
political acts of framing and ratification that a judge ever may invalidate the action of a contemporary political majority.
On the other hand, some degree of flexibility in identifying the
framers' relevant intent is as normatively desirable as it is descriptively
necessary for constructivist coherence theory to be plausible as a
theory of our actual practice. Two modern bodies of law - one
involving freedom of speech, the other concerning equal protection exemplify the moral, political, and legal intuition that the framers'
specific intent should not always be determinative. In neither area
can I claim a historian's expertise. In both, however, there are strong
arguments that our current jurisprudence has diverged very far from
the framers' specific historical intentions. 268 Consider the competing
considerations in the following two cases.
Suppose that the framers of the first amendment specifically intended "at most .
7 (1966) (asserting that suppression of speech is an affront to the dignity of man in that it
frustrates attainment of individual self-fulfillment).
1987]
CONSTRUCTIVIST COHERENCE
1257
intentions of persons long dead, who could have foreseen the future
only imperfectly at best, have a controlling effect? My own strong
intuitive judgment, which is entirely consistent with Supreme Court
precedent, 2 72 is that the specific intent should not be determinative.
Similar intuitions are likely to be felt in the equal protection area.
The most monumental case of the modern era, Brown v. Board of
Education,2 73 evokes an especially confident judgment: even if there
were substantial evidence that most of the framers of the fourteenth
amendment had, upon reflection, concluded that segregated schools
did not offend the equal protection clause, that judgment should not
be dispositive in light of current moral sensibilities, the law's intervening line of growth and development, and the general intent of the
framers to establish a legal regime guaranteeing to all races the equal
274
protection of the laws.
272 See, e.g., Cohen v. California, 403 U.S. 15 (1971) (holding that first amendment protects
the use of a vulgar expletive in communication of political protest); West Va. State Bd. of Educ.
v. Barnette, 319 U.S. 624 (1943) (holding that the first amendment prohibits compulsion of
school children to salute the flag).
273 347 U.S. 483 (1954).
274 For a fuller discussion of Brown and the various kinds of constitutional argument as they
bore on the decision in that case, see pages 1278-82 below.
275 The inconsistency can and should stop short of incoherence or self-contradiction.
See
1258
[Vol. IO0:ii8 9
19871
CONSTRUCTIVIST COHERENCE
1259
281 See, e.g., Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 555-57 (1985)
(overruling Nat'l League of Cities v. Usery, 426 U.S. 833 (1976)).
282 Cf. Schauer, supra note 17, at 811-12 (asserting that the authoritativeness of the constitutional text, as a text, requires attention to textual purposes distinct from the "psychological
conditions" of the Constitution's authors). See generally Fuller, Positivism and Fidelity to Law
A Reply to Professor Hart, 71 HARV. L. REv. 630 (1958).
283 See Murphy, An Ordering of Constitutional Values, 53 SO. CAL. L. REV. 703, 754-57
(i98o) (arguing that constitutional amendments could be unconstitutional if incompatible with
the basic scheme of values to which the Constitution is best viewed as committed). Professor
126o
[Vol. l~oo:1189
theory permitting this result would not only depart from what I take
to be the understanding of most constitutional lawyers; it would accord
too little weight to the claim that the Constitution, whatever else it
may be, is positive law.
D. Arguments from Precedent
Arguments of precedent fit comfortably into the constructivist coherence framework largely because of their theoretical and valueinfused nature. An account of the holdings of a string of cases states
a theory of their meaning that, like any theory, will be value-infused. 284 Moreover, like any legal theory proposed to guide the decision of future cases, a theory explaining prior decisions will be
subject to evaluation under joint and sometimes competing criteria of
descriptive adequacy and normative attractiveness. A concern for
moral and policy values is therefore built into arguments from precedent, and explicit attention to value arguments frequently will have
28 5
a spillover effect into the precedential category.
With precedential arguments exhibiting this substantial degree of
indeterminacy and adaptability, the workability of constructivist coherence theory is clear. The theory holds that plausible interpretations
of the cases should be considered in light of such factors as fidelity to
textual language, the intent of the framers, and considerations of
constitutional theory and of morality and social policy. Value arguments are likely to be especially important. But reciprocal influence
is also possible; the most persuasive reading of the precedents can and
should affect the assessment of arguments based on text, historical
286
intent, constitutional theory, and even moral and policy values.
In order to be persuasive, constructivist coherence theory's interpretation of the role of precedent requires a normative argument that
it is at least reasonable for arguments from precedent to get precisely
as much weight as they do - not less and not more. From the
perspective of those who regard text and the framer's intent as the
sole legitimate authorities, precedent's influence may seem too large.
This is particularly true insofar as arguments of precedent are permitted to affect arguments from text, as when "lawyers' meanings"
7
are allowed to substitute for plain language or historical meanings 28
Dworkin seems occasionally to have flirted with, but apparently rejected, a similar position.
See infra notes 403-o5 and accompanying text.
284 See supra notes 59-65 and accompanying text.
285 If this argument has a modernistic cast, the subtantial malleability of precedent has long
been recognized. See, e.g., E. LEVI, supra note 58, at 1-4; K. LLEWELLYN, THE COMMON
LAW TRADITION 75-91 (ig6o); Radin, Case Law and Stare Decisis: Concerning Prajudizienrecht
in Amerika, 33 COLUM. L. REv. 199 (i933).
286 See infra note 296 and accompanying text.
287 See supra p. 1238.
1987]
CONSTRUCTIVIST COHERENCE
1261
In Reynolds, Justice
Harlan made a powerful argument in dissent that the Court's decision was unjustifiable because
the framers of the fourteenth amendment had no specific intent to create voting rights. See id.
at 59o-9i, 607-08 (Harlan, J., dissenting). Once a long line of decisions has grown up, however,
the existence of other cases proceeding on the assumption that the framers' intent should be
understood as encompassing voting rights provides a reason for understanding their intent in
just that way. Justice Harlan's historical argument has not been repeated in any recent case.
289 See Monaghan, Court Opinions, supra note 56, at 7; Monaghan, supra note 30, at 38283, 387-91.
1262
equality holdings in the sex-marriage-children area," Monaghan, supra note 30, at 382.
The
implications of this cryptic assertion are somewhat less than clear. Yet Monaghan seems to
contemplate a kind of crabbed construction of the underlying decisions' precedential authority
in future cases - a style of analysis generative of unpredictability and arbitrariness, and, within
the context of our legal system, one that is in tension with Monaghan's own concession that
arguments from original intent "cannot now be recognized as the only legitimate mode of
constitutional reasoning." Monaghan, supra note 30, at 382.
292 198 U.S. 45 (1905).
293 Whether it actually did so is a question that I mean to leave open. "[Wihile nearly 200
regulations were struck down during the Lochner era, most challenged laws withstood attack."
G. GUNTHER, CONSTITUTIONAL LAW 453-54 (iIth ed. 1985).
1987]
CONSTRUCTIVIST COHERENCE
1263
terpreter to review any balance that she has struck within other
categories.
Just as moral and political considerations affect the way that other
kinds of argument get weighed in constitutional interpretation, there
is a reciprocal traffic: value arguments themselves may be influenced
by other factors in constitutional analysis. From whatever source they
are derived, particular value judgments must be reconsidered if they
are incompatible with the most plausible interpretations of constitutional text, historical intent, constitutional theory, or precedent.
Checking particularistic judgments against other indicators of moral
and political insight can only enrich the process of decision. Even as
moral and political judgments, conclusions ought to be suspect if
dramatically divergent from the traditions embodied in law. 296 And
the demand for reconsideration and possible adjustment of moral or
political views is especially apt when moral or political arguments
function within a legal calculus. Constructivist coherence theory is
concerned with the use of value arguments within the practice of
constitutional interpretation, not in independent moral or political
reasoning.
Although it defines the framework within which value arguments
operate, constructivist coherence theory does not provide extensive
criteria for identifying the best arguments of value or even the foundational theories in which those arguments would be grounded. The
theory's principal significance, beyond specifying how the various categories of argument do and ought to interact, lies in its identification
of a kind of filter: when value arguments function as constituent
elements of constitutional reasoning, a constitutional interpreter must
be able to make at least the implicit claim that the values to which
297
she appeals are in some sense those of the constitutional community.
296 The idea that tentative moral judgments or intuitions are subject to reconsideration and
adjustment - although typically in a two-way traffic with more general moral virtues or
principles - is a familiar one in moral and political philosophy. See supra note 230.
297 This demand is not a vacuous one. But neither should its significance be overstated.
Although I assume that some sources of value would prove unworkable in practice and therefore
need to be abandoned, I would be surprised if this happened with respect to relatively familiar,
even if controversial, moral or political theories. Moreover, the identification of values that are
in some sense those of the constitutional community is an interpretive enterprise that is deeply
contestable. Professor Robert Cover, for example, has argued forcefully and provocatively that
our nation is home to a disparate array of groups, each of which defines itself through a different
conception of normative order through which it perceives and into which it seeks to incorporate
institutions and events. See Cover, Violence and the Word, 95 YALE L.J. 16oi (1986); Cover,
supra note 26. This fact - for I take it to be such - does not defeat the enterprise of
identifying the values of the constitutional community. For the community's values, as they
typically are and must be assumed to be for purposes of constitutional argument, do not represent
the sum of some set of observable behaviors or even responses to a public opinion survey.
Rather, this concept must express the familiar interpretive efforts of judges, critics, and social
philosophers to articulate the principles and commitments that best explain particular traditions,
1264
[Vol. 100:1189
19871
CONSTRUCTIVIST COHERENCE
1265
302 See C. BLACK, THE PEOPLE AND THE COURT 6 (196o) (arguing that if the notion that
the Constitution is law is rejected, "then the American doctrine of judicial review has no
theoretical justification").
303 The thesis that moral and legal reasoning are inseparable in all cases might be asserted
on any of three bases. One is a thesis that the law is totally indeterminate and that legal
reasoning therefore reduces to moral and political reasoning about what is the best thing to do
in any particular case. Cf. Kennedy, Legal Education as Trainingfor Hierarchy, in THE
POLITICS OF LAW 40, 47 (D. Kairys ed. 1982) ("There is never a 'correct legal conclusion' that
is other than the correct ethical and political solution to legal issues."). This view is inconsistent
with most of the argument in this essay. The assumption reflected here is that the norms and
shared understandings obtaining within the practice of constitutional interpretation at a given
time will sufficiently cabin the range of plausible interpretations so that any total indeterminacy
thesis - if it is to be stronger than the Humpty Dumpty claim that anyone can use words
however she chooses - must be rejected. A second basis for the conclusion that legal and
moral reasoning are universally inseparable would be that our constitutional system is sufficiently
infused with natural law that it just happens that the legally best answer in every case will be
identical with the morally best answer. I regard this vision of the American Constitution as
impossible to sustain. The constitutional status of slavery once provided convincing proof. See
infra notes 399-405 and accompanying text. The Constitution's relative indifference to issues
of poverty and distributive justice continues to reflect, for all but the most extreme Lockean,
see e.g., R. NOZICK, ANARCHY, STATE, AND UTOPIA (1974), the misguidedness of a panglossian
outlook. A third possible argument would be that, because the decision of every case requires
judgments that are "value-infused," politics or morality enters into every decision and that the
"partial autonomy" thesis therefore collapses. Although the premise of this argument is a valid
one, the asserted conclusion does not follow. The "partial autonomy" thesis does not deny "value
infusionism"; it implies only that "the best legal argument" can sometimes be different from "the
best moral argument" and, thus, that legal reasoning is not identical with moral reasoning.
304 But see Perry, supra note 17, at 564, 583 (arguing that judges do and should exercise a
"prophetic function").
305 See generally Fried, The Artificial Reason of the Law or: What Lawyers Know, 6o TEX.
L. REv. 35 (1981).
1266
[Vol. Ioo:II89
future").
o30
6o U.S. (xg How.) 393 (1856).
309 i98 U.S. 45 (195o).
310 See Rehnquist, The Notion of a Living Constitution, 54 TEx. L. REv. 693, 700-04
(1976).
CONSTRUCTIVIST COHERENCE
1987]
1267
Important precursors to that decision included Baker v. Carr, 369 U.S. i86 (1962), and Wesberry
v. Sanders, 376 U.S.
316
317
(1964).
1268
[Vol. 1oo:i1i8 9
abjure reliance on value arguments. Democracy is a deeply contestable concept. Certainly our democracy, with its assumption that contemporary majorities are subject to higher law, has acquiesced in and
even depended upon a constitutional practice in which judges rely
heavily on value arguments in rendering their decisions. 3 18 Our familiar practice of judicial review, including its reliance on value arguments, enhances rather than thwarts the distinctively American
system of political democracy in which the people have trusted the
courts, and especially the Supreme Court, to function as a check and
a break - to signal that certain things that currently seem desirable
or expedient ought not to be done because they are incompatible with
the moral or political principles on which the constitutional community
is built.
The argument from practical necessity, which is closely related to
the argument from competence, follows from the premise of the indeterminacy of other legal materials in hard constitutional cases.
Where other sources of law would permit more than one result, it is
only reasonable for a decisionmaker to employ moral and political
criteria as grounds for preference. Often this result may occur through
arguments or conclusions that could be described as "merely" valueinfused. But if values, explicitly or implicitly, must enter into the
process of decision, candor demands that the truth be recognized. 3 19
Moreover, by flushing moral judgments out into the open, constructivist coherence theory invites judges to confront the extent to which
their views really are infused by moral and political values and to
assess arguments and evidence that their beliefs may be mistaken. A
moral judgment not only can but frequently will be too uncertain to
justify forced efforts to achieve coherence in favor of a particular
decision. This important truth is as likely to be grasped by judges
who are self-conscious about their reliance on value arguments as by
those whose reliance is not so much less as it is less self-conscious.
VI.
general or public good, not the narrower interests of partisan groups, and that courts act
consistently with the Madisonian understanding of democracy when conducting deliberations
about the public good).
319 See Shapiro, supra note 59.
19871
CONSTRUCTIVIST COHERENCE
1269
320
1270
[Vol. 1oo:1189
by the Supreme Court in a rhetorical mode that suggested a constitutional ideal, and arguably a requirement, of color blindness. 322 As
Professor Randall Kennedy has noted: "In the forties, fifties and early
sixties, against the backdrop of laws that used racial distinctions to
exclude Negroes . . .it seemed that racial subjugation could be overcome by mandating the application of race-blind law. '3 23 By the late
sixties and early seventies, however, blacks continued to lag badly
behind whites in the most critical indicators of educational and economic success. 3 24 The impulse to adopt programs of affirmative action
arose during this period.32 5 In order to achieve a tangible closing of
the economic and educational gap between blacks and whites, a variety of institutions, both public and private, undertook affirmative
action initiatives that granted explicit preferences to blacks and other
minorities.
The Supreme Court first confronted a challenge to the legality of
affirmative action in 1978, in Regents of the University of California
v. Bakke. 32 6 Specifically at issue in Bakke was a system of minority
admissions preferences, including a racial quota, administered by the
medical school of the University of California at Davis. By a vote of
5-4, the Supreme Court held the racial quota impermissible. 3 27 Also
by a vote of 5-4, however, the Court concluded that noninvidious
racial preferences could be upheld if closely enough related to a sufficiently important governmental interest. 3 28 The Court was badly
fractured. Only Justice Powell, who announced the judgment in an
opinion that was not joined in full by any other Justice, concurred in
both aspects of the holding. 32 9 But my interest in Bakke is less in the
specific conclusions reached by the participating justices than in the
322 See Kennedy, Persuasionand Distrust: A Comment on the Affirmative Action Debate, 99
HARv. L. REv. 1327, 1335-36 & n.32 (1986); Reynolds, Individualism v. Group Rights: The
Legacy of Brown, 93 YALE L.J. 995, 998-1000 (1984).
323 Kennedy, supra note 322, at 1335.
324 See Regents of the Univ. of Calif. v. Bakke, 438 U.S. 265, 395-96 (1978) (opinion of
Marshall, J.); Fallon & Weiler, Firefighters v. Stotts: Conflicting Models of Racial Justice, 1984
SuP. CT. REV. 1, 33.
325 See Fallon & Weiler, supra note 324, at 33-34.
326 438 U.S. 265 (1978). The Court previously had heard arguments about the constitutionality of affirmative action in DeFunis v. Odegaard, 416 U.S. 312 (I974), but held the case to
important and articulated purpose," as long as they do not "stigmatize[ ] any group or... single[
] out those least well represented in the political process to bear the brunt of [the] program."
Id. at 361 (opinion of Brennan, White, Marshall, and Blackmun, JJ.). But Justice Powell, who
provided the crucial fifth vote for the result, found even noninvidious racial classifications to
be permissible only when "precisely tailored," id. at 299 (opinion of Powell, J.), or even
necessary, id. at 320, to serve a "compelling government interest," id. at 299.
329 See id. at 271-72.
1987]
CONSTRUCTIVIST COHERENCE
1271
preferences fit into the background web of moral concerns and concepts in which "equal protection" is rooted. Also potentially relevant
is how judicial precedents have previously construed "equal protection" as a legal concept. In short, arguments from text blend at the
outset with arguments concerning the framers' intent, moral and political values, and precedent.
Within the category of the framers' intent, the question in Bakke
was whether the thirty-ninth Congress meant to prohibit race-based
preferences designed to rectify past societal discrimination. 3 3 1 This
question is vastly less straightforward than it appears. It is no accident that the legislative history of the fourteenth amendment does not
speak directly to the permissibility of affirmative action. The possibility of affirmative action preferences that were not closely linked to.
the abolition of slavery was virtually unthinkable at the time of the
amendment's enactment. Reconstruction congresses did create the
330 See supra pp. 12o5-o6.
331 One arguable account is that the framers intended to extend the scope of the equal
protection guarantee only to a closed list of rights - not including education - deemed to be
"fundamental" in 1868. See R. BERGER, supra note 39, at 133, 168-92. But this approach has
been foreclosed, at least for anyone accepting the constitutional soundness of Brown v. Board
of Educ., 347 U.S. 483 (I954).
1272
[Vol. lOO:1189
Freedmen's Bureau to assist freed blacks in attaining economic independence. 332 But no thought was given to whether, years after slavery
had been abolished, affirmative action would be permissible not to
compensate individuals for wrongs done to them individually, but to
disestablish patterns of societal discrimination that caused disadvantage to blacks as a group. With respect to this question, the framers'
intent might simply be pronounced inconclusive. Alternatively, the
inquiry could be conceived more abstractly: what stance toward affirmative action follows from the moral and political theory that best
explains the relevant historical evidence of the framers' actual intent?
If the question is formulated in this way, however, it has no purely
historical answer. The required judgment, which is interpretive, has
333
an irreducibly normative dimension.
Precedent and prevalent moral and political judgments would have
helped in Bakke to frame at least three plausible specifications of the
framers' abstract intent. From one view, the framers' central concern
was to forbid discriminations based on race. 33 4 Another specification
of the framers' intent would be less absolute. According to this account, the framers' abstract intent, growing out of their sense of the
wrongfulness of slavery and the attendant subjugation of blacks to an
inferior status, was to render strongly suspect any discrimination based
on race or other qualities that commonly are the basis for invidious
discriminations. 3 35 Finally, the framers' intent might be characterized
directly in terms of invidious motivation. With Reconstruction congresses having created the Freedmen's Bureau to assist newly freed
blacks, it is arguable that the framers would have meant to ban only
those discriminations that resulted from animosity or indifference to
3 36
the disadvantaged group.
Issues of constitutional theory were also present in Bakke. It is
possible to imagine a constitutional theory, developed around a particular conception of equality, that would forbid affirmative action.
In Fullilove v. Klutznick,3 37 for example, Justice Stevens propounded
a theory that linked the equal protection clause with the constitutional
prohibition against titles of nobility. 338 In conjunction, he thought,
332 See Bakke, 438 U.S. at 391, 397-98 (opinion of Marshall, J.).
333 See supra pp. 1212-13.
334 This is the account presented by Justice Stewart in his dissent in Fullilove v. Klutznick,
448 U.S. 448, 523-26 (ig8o) (Stewart, J., dissenting).
33s Cf. Bakke, 438 U.S. at 291-99 (opinion of Powell, J.) (requiring a "compelling" justification for any race-based governmental classification and tracing this requirement, in part, to
1987]
CONSTRUCTIVIST COHERENCE
1273
1274
[Vol. IOO:II8 9
1987]
CONSTRUCTIVIST COHERENCE
1275
352 See, e.g., Bakke, 438 U.S. at 299 (opinion of Powell, J.) (formulating this standard).
Echoing another familiar formulation from the case law, I would also have concluded that an
affirmative action program, in order to be upheld, must be necessary to serve, or at least be
very closely tailored to, the justifying interest. See, e.g., id. at 299, 320.
353 See supra note 332 and accompanying text.
354 See Fallon & Weiler, supra note 324, at 30 (defining rectification as "the elimination from
economic and social institutions of the presumptive incidents of past discrimination, including
racial skewings in the distribution of wealth, jobs, power, and opportunities").
355 See id. at 30-37.
356 See, e.g., Griggs v. Duke Power Co., 401 U.S. 424, 432 (197) (interpreting racial
discrimination under title VII of the Civil Rights Act of 1964 as including "the consequences of
employment practices," even when discriminatory intent is absent (emphasis in original)).
1276
[Vol. ioo:ii89
1987]
1277
Justices Brennan, White, Marshall, and Blackmun reached a different alignment of the various relevant arguments. In a jointly authored opinion, these Justices determined that racial classifications used
to remedy past societal discrimination should be upheld so long as
they did not "stigmatize[ ] any group or ... single[ ] out those least
well represented in the political process to bear the brunt" of the
burden. 36 2 Their analysis relied heavily on value arguments and an
interpretation of the precedents as forbidding only invidious discrimination. 3 63 Yet the joint opinion also takes a clear if implicit stance
with respect to arguments from text and of constitutional theory. In
addition, Justice Marshall, in his separate opinion, made explicit ar364
guments concerning the framers' intent.
The four other Justices who participated in Bakke never reached
the constitutional issue, because they concluded that the affirmative
action program involved in the case was forbidden by a federal statute. 36 5 Nevertheless, the possibility of yet another alignment was
demonstrated by the Supreme Court's next major affirmative action
case, Fullilove v. Klutznick, 3 66 in which Justices Stewart and Rehnquist achieved coherence among the factors in arguing that the Con36 7
stitution forbids all race-based classifications.
In light of the variety of conclusions that can be reached within
constructivist coherence theory, 3 68 the theory's value inheres in its
capacity to clarify and to systematize norms to which most constitutional lawyers have tacitly conformed all along. However adequate
our tacit knowledge of the norms of our practice generally may be,
moments of confusion and uncertainty can arise, and it is useful to
have an articulated statement of the standards to which our practice
commits us.
A final word about Bakke may also be useful. Because I have
emphasized the connectedness among the five categories of constitutional argument, it might appear that constructivist coherence theory
362 Id. at 361 (opinion of Brennan, White, Marshall, and Blackmun, JJ.). These four Justices
would also have required the remedial program to be substantially related to its justifying
purpose. Id. at 359.
363 See id. at 358-62 (discussing value arguments); id. at 355-56 (citing precedent).
364 See id. at 391, 397-98 (opinion of Marshall, J.).
36- See Bakke, 438 U.S. at 412-21 (opinion of Stevens, J., joined by Stewart and Rehnquist,
JJ., and Burger, C.J.).
366 448 U.S. 448 (1980).
367 See id. at 522-27 (Stewart, J., joined by Rehnquist, J., dissenting).
368 Although constructivist coherence theory entails no unique choice among these contending
lines of argument, some conclusions may certainly be better than others; indeed, one judgment
may be right and the others wrong. But the normative commitments needed to resolve such
substantive constitutional issues as those surrounding affirmative action go far beyond the
commensurability problem of how the different kinds of constitutional argument fit together in
a single constitutional calculus.
1278
[Vol. ioo:i i8 9
pp. 1282-85.
370 Any honest originalist must recognize a deep problem about the role of precedent in our
constitutional practice. See supra pp. 1202-04, 1213.
371 See Bakke, 438 U.S. at 289-90 (opinion of Powell, J.).
372 See id. at 291-93.
373See id. at 290-91, 294-95, 300-05.
374See id. at 290-91, 298-99.
375The one exception involves the joint opinion's treatment of the framers' intent. The
jointly authored opinion of Justices Brennan, White, Marshall, and Blackmun merges its discussion of the framers' intent with arguments of precedent and argues that the equal protection
clause has not been construed to require absolute color blindness. See id. at 355-56 (opinion
of Brennan, White, Marshall, and Blackmun, JJ.). In a separate opinion, Justice Marshall
addressed the issue of the framers' intent somewhat more directly. See id. at 396-98 (opinion
of Marshall, J.).
376 347 U.S. 483 (1954).
377The moral and political reasoning was evident on the face of the Court's opinion:
education, because of its formative effect on children, "is perhaps the most important function
of state and local governments." Brown, 347 U.S. at 493. To separate children based on their
race "generates a feeling of inferiority [in black children] . . .that may affect their hearts and
minds in a way unlikely ever to be undone." Id. at 494. And to stigmatize or otherwise harm
children in this way is wrong. See Black, The Lawfulness of the Segregation Decisions, 69
YALE L.J. 421, 425-26 (I96O).
19871
CONSTRUCTIVIST COHERENCE
1279
128o
[Vol. lOO:118 9
compatible with strongly determined arguments from text and of constitutional theory and with a tenable construction of the framers'
intent. In Brown, the value arguments that infused the categories of
arguments from text and of constititional theory very arguably would
have justified an overruling.
The main sticking point in Brown, and the principal source of
continuing interest in the case for constitutional theory, involved the
framers' intent. As noted above, it is plausible to view Congress as
having had no specific intent to require school desegregation when it
enacted the fourteenth amendment. 383 In his opinion for the Court
in Brown, Chief Justice Warren disposed of the intent issue somewhat
summarily. The historical sources regarding the framers' intent "cast
some light," he wrote, but "[a]t best, they are inconclusive. '384 Although compatible in principle with the constructivist coherence methodology, this treatment of the framers' intent is too cryptic to command much credence. It appears, however, that there are at least
two bases on which the framers' intent might be viewed as capable
of cohering with the result indicated by the arguments from other
constitutional categories. First, the framers' intent could be specified
in abstract terms as, for example, the intent to embed within the
Constitution the principle that blacks may not be unreasonably and
substantially disadvantaged by state legislation, 3 85 or, more generally,
that no citizen may be denied equal concern and respect by state
3 86
authorities.
An alternative theory, which would equally justify and perhaps
better explain the Court's result, was articulated by Professor Alexander Bickel, who, as a Supreme Court law clerk at the time that
Brown was argued, wrote a memorandum on the legislative history
of the fourteenth amendment that Justice Frankfurter circulated to
believes that government should treat people as equals, and that this means not treating
them differently with respect to their fundamental interests. He believes that the clause
he votes for would be violated by criminal laws providing different penalties for black
[sic] and whites guilty of the same crime, for example, because he believes that liability
to punishment touches a fundamental interest. But he also believes that separate and
unequal public schools would not violate the clause, because he does not consider education to be a fundamental interest. Once again we can distinguish an abstract and
concrete formulation of his intention.
19871
CONSTRUCTIVIST COHERENCE
1281
the entire Court. 38 7 After hearing oral argument in Brown and taking
the case under submission, the Supreme Court set the case for reargument and requested special briefing on two issues. The Court
queried first whether Congress and the state legislatures contemplated
that the fourteenth amendment would abolish segregation in the public
schools. 388 The Court then asked, as a separate question, whether
the framers anticipated that, even if segregation were not abolished
immediately, the courts would have the power under future conditions
to hold segregation unconstitutional. 38 9 In response to this latter question, Bickel concluded that it was at least plausible that the framers
deliberately chose language capable of supporting a gradual expansion
390
of the jurisprudence of equality.
With the radical Republicans in the thirty-ninth Congress preferring sweeping and explicit guarantees for blacks, and the moderates
preferring an approach that was socially and politically more conservative, Bickel suggested that the general and expansive language of
the equal protection clause may have reflected a compromise - one
that permitted moderate Republicans immediate defense against
"alarms raised by the opposition [against the abolition of segregation],
but which at the same time was sufficiently elastic to permit reasonable future advances." '39 1 Bickel's hypothesis was only a hypothesis
- sufficiently doubtful, he judged, so that if the equal protection
clause were part of a statute, the supporting evidence would not be
strong enough to sustain his theory. 392 But in light of the other factors
bearing on the decision in Brown, Bickel judged his hypothesis plausible enough to justify the conclusion that the framers' intent, for the
Court's purposes, fairly could be termed "'inconclusive.'
3 93
It was
1282
[Vol. xoo:1189
Although Bickel's argument provides general support for the approach to questions of the framers' intent that is modelled by constructivist coherence theory, his defense of Brown illuminates an issue of
general importance within our constitutional interpretive practice: the
extent to which we can properly allow arguments within one category
of constitutional argument to affect conclusions within another. To
say that interdependences among the categories exist, and that our
practice encourages their exploitation, is not to say how far the influence of arguments within one category should be allowed to extend
into another. In Brown, the Court seems to have stayed well within
the bounds of permissibility in holding evidence of the framers' intent
to be too "inconclusive" to determine the issue before it, despite the
prima facie plausibility of the conclusion that the framers had a
relatively clear specific intent beyond which they meant to authorize
no judicial invalidations of settled social practices. But the problem
of how far conclusions may appropriately be adjusted in order to
achieve coherence is a general one; it would be possible to cite instances of more doubtful adjustments of conclusions within one category to accord with the results prescribed by others. It is in part
because Brown exposes this general issue that it deserves to be treated
as a paradigm case.
C. The ConstitutionalLaw of Slavery
Constructivist coherence theory is largely a theory about the potential for attaining coherence that results from connectedness among
the categories of constitutional argument. But the theory is also one
about hierarchy and constraint. When arguments from text and from
the framers' intent prove resistant to accommodation, their hierarchical authority demands recognition. And while the range of permissible
accommodations is broad, the hierarchical ordering of categories of
argument presumes that there are limits. Some arguments, some
attempts to reach equilibrium, transgress the outward bounds of colorable argument.
Within any particular era, debates over the permissible limits of
accommodation are likely to focus on a relative handful of cases or
issues. A contemporary example, to which I have alluded already,
394
Id. at 64-65.
1987]
CONSTRUCTIVIST COHERENCE
1283
No Person held to Service or Labour in one State, under the Laws thereof, escaping into
another, shall, in Consequence of any Law or Regulation therein, be discharged from
such Service or Labor, but shall be delivered up on Claim of the Party to whom such
Service or Labour may be due.
U.S. CONST. art. IV, 2, cl.3; accord U.S. CONST. art I, 2, cl.3 (distinguishing between
"free Persons" and "other Persons" in providing for apportionment of states' representation in
the House of Representatives).
401 See M. FARRAND, THE FRAMING OF THE CONSTITUTION OF THE UNITED
STATES 14752 (1913); E. MORGAN, THE BIRTH OF THE REPUBLIC 1763-1789, at 141-42 (1956).
1284
[Vol. lOO:1189
403 Dworkin, The Law of the Slave Catchers, 1975 TIMES LITERARY SUPPLEMENT (London)
1437 (reviewing R. COVER, JUSTICE ACCUSED: ANTISLAVERY AND THE JUDICIAL PROCESS
(1975)).
404 Dworkin discusses the general "theory of mistakes" in TAKING RIGHTS SERIOUSLY, supra
note 37, at 121-23.
405 Dworkin, supra note 403, at 1437. Dworkin's cryptic suggestion to this effect does not
seem consistent with his nearly contemporaneous argument, published in 1975 in the Harvard
Law Review and later reprinted in Taking Rights Seriously, cited in note 37 above, that a theory
of mistakes, which is most important with respect to judicial precedents, must be much more
modest in its approach to constitutional provisions. In dealing with constitutional provisions,
Dworkin argued that a theory of mistakes must distinguish between a provision's "gravitational
force" - its capacity, because of the weight of the principles that support it, to influence the
results in cases in which it is not literally applicable - and its "enactment force," which defines
the minimal reach that it must possess in order for its status as positive law to be respected.
See R. DWORKIN, TAKING RIGHTS SERIOUSLY, supra note 37, at 121-23. Lack of fit with the
best political theory explaining the Constitution as a whole could provide a reason for denying
much gravitational force to the provisions countenancing slavery, but it would seem a minimal
entailment of their enactment force, which a conscientious judge would be required to respect,
that slavery could not simply be unconstitutional. For an example of a constitutional theory
that would allow judicial rejection even of clear constitutional text, see Murphy, cited in note
283 above, at 754-56, who argues that constitutional amendments would be unconstitutional if
they departed too far from the moral and political understandings embedded in the basic
document and its surrounding traditions.
1987]
CONSTRUCTIVIST COHERENCE
1285
VII. CONCLUSION
Constitutional scholarship has given remarkably little attention to
the commensurability problem - the problem of how different kinds
of constitutional argument are appropriately combined and weighed
against each other within our constitutional practice. The significance
406
Act of Feb.
12,
1793, ch. 7, i Stat. 302; An Act to amend and supplementary to the Act
entitled 'An Act respecting Fugitives from Justice, and Persons escaping from the Service of
I84os and 185os, however, the Supreme Court upheld the constitutionality of the Fugitive Slave
Acts on at least four occasions. See Ableman v. Booth, 62 U.S. (21 How.) 5o6 (I859); Moore
v. Illinois, 55 U.S. (14 How.) 13 (1855); Jones v. Van Zandt, 46 U.S. (5 How.) 215 (1847); Prigg
v. Pennsylvania, 41 U.S. (6
questions raised by the Fugitive Slave Acts, see R. COVER, cited in note 403 above, at 159-91;
Dworkin, cited in note 403 above, at 1437.
1286
[Vol. lOO:118 9
of this problem is self-evident. Judges and lawyers in our constitutional practice commonly rely on at least five different kinds of arguments - arguments from text, the framers' intent, constitutional
theory, precedent, and moral and policy values. How are these different kinds of arguments related to each other? Are all entitled to
influence in every case? Which take precedence in cases of conflict?
Questions such as these need answers if we are to understand what
we do and ought to do as constitutional lawyers.
Little previous writing has addressed the commensurability problem directly, and this essay has thus had several purposes. The first
has been simply to frame the relevant issues more clearly than has
been done before. The second has been to demonstrate that the
responses of leading constitutional theories are inadequate. The third
has been to commend a methodological stance toward the problem
that blends descriptive and normative analysis. And the fourth has
been to identify the solution that is implicit in our actual practice.
The solution, which I call constructivist coherence theory, has two
aspects. The first relies on the central conceptual idea of constructivist
coherence: the notion that the various types of argument function not
autonomously but interactively. The constructivist coherence approach assumes that, even when a tentative assessment of arguments
within various categories suggests a conflict among prescribed results,
the balance of competing arguments frequently can be reconsidered
in a successful effort to achieve a uniform prescription. Sometimes,
however, the strongest arguments within the different categories will
point irreversibly to different conclusions. In such cases, the theory's
second aspect comes into play. The implicit norms of our constitutional practice, I have argued, require that the claims of the different
kinds of arguments be ranked hierarchically.
Constructivist coherence theory, which undertakes to identify the
implicit rule-like structure that underlies our constitutional practice,
provides a framework for those puzzled about how different kinds of
constitutional argument fit together. Others, perhaps less puzzled,
may produce different answers to the questions that I have raised. I
do not, in conclusion, wish to undermine my own constructive argument. But I do want to emphasize the larger significance of the
questions advanced by this essay. Though long ignored, the commensurability problem demands the thoughtful attention of anyone who
wishes to engage in constitutional argument.